Loop resistance measuring method, system and device

Abstract

The invention discloses a loop resistance measuring method, system and device, and relates to the technical field of electrical measurement. The method comprises the following steps of: performing loop data acquisition at each interface of the disconnecting switch adopting a three-channel wiring mode, injecting excitation current by an excitation source, and acquiring three-channel voltage original signals and initial sampling flow by a differential isolation amplifier to obtain the initial sampling flow; time sequence identification is carried out on the initial sampling flow to complete sample starting and ending point identification, inter-channel time alignment is completed, and a synchronous voltage sample set is output. According to the invention, by constructing a complete data chain covering high-precision synchronous sampling, adaptive signal processing, phase compensation calculation and intelligent statistical diagnosis, the problems of low measurement precision and insufficient anomaly identification capability caused by poor synchronism, phase error and lack of deep analysis in a traditional method are solved; the accuracy and anti-interference performance of loop resistance measurement and the intelligent level of state diagnosis are improved.

Description

Technical Field

[0001] This invention belongs to the field of electrical measurement technology, and in particular to a method, system and device for measuring loop resistance. Background Technology

[0002] Currently, there are significant limitations in the technology for measuring the resistance of disconnector circuits. Traditional DC voltage drop methods are susceptible to field thermoelectric potential and electromagnetic interference, leading to data drift and poor repeatability. AC excitation schemes generally ignore the phase asynchrony problem between multiple channels; phase errors introduced by hardware delays and transmission path differences directly reduce the accuracy of resistance calculations. Existing methods have rigid signal processing mechanisms, employing fixed sampling windows and filtering parameters, unable to adaptively adjust based on real-time signal-to-noise ratio, resulting in insufficient robustness under complex operating conditions. Furthermore, most techniques only provide single resistance measurements, lacking the ability to statistically analyze continuous measurement data. This makes it impossible to assess the confidence level of measurement results or automatically identify contact anomalies using quantitative indicators (such as imbalance rate), still relying on manual experience for judgment, which is inefficient and prone to missed detections. Summary of the Invention

[0003] The purpose of this invention is to address the problems existing in the prior art by providing a method, system, and apparatus for measuring loop resistance.

[0004] The technical solution to achieve the objective of this invention is as follows: On one hand, a method for measuring loop resistance is provided, the method comprising the following steps:

[0005] Step 1: Loop data is acquired at each interface of the disconnector using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream.

[0006] Step 2: Implement timing identification on the initial sampling stream to complete the identification of the sample start and end points, complete the time alignment between channels, and output a synchronous voltage sample set;

[0007] Step 3: Construct a set of endpoint voltage amplitude and phase features from the set of synchronous voltage samples;

[0008] Step 4: Based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling current, the voltage difference after phase compensation is divided by the corresponding excitation current amplitude to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment, and the steady-state estimation is performed on the instantaneous estimation sequence to generate a set of loop resistance estimated values.

[0009] Step 5: Calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and judge contact abnormalities according to the preset judgment threshold and confidence rules. Output a list of contact abnormalities, including the corresponding interface identifier and measurement timestamp.

[0010] Furthermore, step 2 specifically includes:

[0011] Step 2-1: Filter each sampling window using a programmable digital bandpass filter to remove the DC component;

[0012] Step 2-2: Calculate the short-time energy spectrum window by window after filtering and identify stable periods based on the spectrum stability criterion. Then, construct a set of effective sampling segments by time series for the identified stable periods.

[0013] Steps 2-3: Extract the instantaneous voltage values ​​of each channel from the set of effective sampling segments according to time alignment to form a candidate synchronization sample set, and select the best to generate the synchronization voltage sample set.

[0014] Furthermore, step 2-2, which involves calculating the short-time energy spectrum window by window after filtering and identifying stable periods based on the spectral stability criterion, specifically includes:

[0015] For each sampling window, a stability score is calculated based on the instantaneous signal-to-noise ratio; the stability score is obtained by combining the energy concentration and the variability rate of the spectral shape over time within each sampling window.

[0016] A threshold is set using stability scores, and windows that pass the stability test are selected to form a subset of stable windows;

[0017] A stable window subset is used as the source of the candidate synchronization sample set instead of the effective sampling segment set. The temporal distribution and spectral characteristics of the stable window subset constitute the effective sampling segment set.

[0018] Furthermore, the set constituting the effective sampling segment also includes:

[0019] The adaptive window length parameter is automatically determined based on the instantaneous signal-to-noise ratio and spectral distribution of samples in the candidate synchronization sample set;

[0020] The adaptive window length parameter is used to adjust the temporal distribution of the sampling window, and the adaptive window length parameter is passed to the passband parameter of the bandpass filter to adjust the spectral characteristics of the filter bandwidth.

[0021] Under low instantaneous signal-to-noise ratio conditions, the sampling window is expanded according to the adaptive window length parameter to enhance statistical stability, while under high instantaneous signal-to-noise ratio conditions, the sampling window is shortened to improve temporal resolution.

[0022] The adaptive window length parameter is recorded in the set of valid sampling segments.

[0023] Furthermore, step 3 specifically includes:

[0024] Step 3-1: For each valid sampling segment within the synchronous voltage sample set, calculate the peak value, root mean square value, and envelope mean value by channel to form an amplitude characterization subset;

[0025] Step 3-2: Phase envelope tracking is performed on each effective sampling segment to obtain the instantaneous phase sequence. Time-domain averaging and phase consistency evaluation are performed on the phase sequences of each channel within the same effective sampling segment to form a phase characterization subset.

[0026] Step 3-3: Merge the amplitude characterization subset and the phase characterization subset, and perform noise weighting to generate the terminal voltage amplitude and phase feature set.

[0027] Furthermore, step 4 specifically includes:

[0028] Step 4-1: For each valid sampling segment in the synchronous voltage sample set, extract the endpoint voltage amplitude and phase feature subsets of the corresponding two channels according to the interface, wherein the endpoint voltage amplitude and phase feature subsets are obtained from the endpoint voltage amplitude and phase feature set.

[0029] Step 4-2: Calculate the instantaneous phase difference between the two channels within the same effective sampling segment based on the amplitude and phase feature subset of the endpoint voltage, and generate phase compensation parameters therefrom, wherein the phase compensation parameters constitute the phase compensation data;

[0030] Step 4-3: Apply phase compensation data to perform phase correction on the terminal voltage amplitude, obtain the phase-compensated terminal voltage amplitude sequence, and record the phase compensation metadata used for correction.

[0031] Step 4-4: Calculate the voltage difference between the two sides after phase compensation according to the interface in each effective sampling segment, and obtain the estimated value of the instantaneous loop resistance of the interface in the effective sampling segment by dividing the voltage difference by the corresponding excitation current amplitude point by point in time, based on the corresponding initial sampling current amplitude in the effective sampling segment.

[0032] Steps 4-5 involve performing steady-state estimation processing on the instantaneous loop resistance estimation sequence to suppress transient noise and outliers. The steady-state estimation processing includes median filtering within the window, outlier removal based on robust statistics, and weighted average calculation. The processing result is the steady-state loop resistance estimation value of the effective sampling segment.

[0033] Steps 4-6: Summarize the estimated steady-state loop resistance values ​​of each effective sampling segment according to the time series to form a set of estimated loop resistance values.

[0034] Furthermore, in step 5, the imbalance rate of each interface is calculated based on the estimated loop resistance values, and contact anomalies are determined according to preset judgment thresholds and confidence rules. Specifically, this includes:

[0035] For each interface, a relative deviation sequence is constructed based on the set of estimated loop resistance values, where the relative deviation sequence is the relative difference between the estimated loop resistance value of the interface and the mean of the set of estimated loop resistance values ​​of the interface;

[0036] Calculate the mean and standard deviation of the biased series, and form the imbalance rate data using the joint judgment criterion of the mean and standard deviation;

[0037] The imbalance rate data is compared with a preset judgment threshold and combined with the confidence score to generate contact anomaly records, thereby constructing a contact anomaly list, which includes the corresponding interface identifier and measurement timestamp.

[0038] Furthermore, the contact anomaly list in step 5 specifically includes:

[0039] Calculate the confidence score based on the consistency of loop resistance estimates across multiple independent valid sampling segments using the same interface;

[0040] A composite judgment index is formed by combining imbalance rate data and confidence score, and the items in the contact anomaly list are assigned priority ranking based on the composite judgment index.

[0041] Priority sorting is used to output priority-based measurement reports from portable terminals or remote monitoring terminals. The measurement reports include the corresponding interface identifier and measurement timestamp.

[0042] On the other hand, a loop resistance measurement system is provided for performing the loop resistance measurement method, the loop resistance measurement system comprising:

[0043] The three-channel acquisition module is used to acquire loop data at each interface of the disconnector switch using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream.

[0044] The signal synchronization module is used to implement timing identification on the initial sampling stream to complete the identification of the start and end points of the samples, and to complete the time alignment between channels and output a set of synchronized voltage samples.

[0045] The endpoint voltage feature extraction module is used to construct an endpoint voltage amplitude and phase feature set from the synchronous voltage sample set;

[0046] The resistance calculation module is used to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling current, by dividing the voltage difference after phase compensation by the corresponding excitation current amplitude, and to perform steady-state estimation on the instantaneous estimation sequence to generate a set of loop resistance estimated values.

[0047] The anomaly detection module is used to calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and to judge contact anomalies according to preset judgment thresholds and confidence rules. It outputs a list of contact anomalies, including the corresponding interface identifier and measurement timestamp.

[0048] On the other hand, a loop resistance measuring device is provided, including an excitation source, a voltage acquisition circuit, a synchronization control circuit, a signal processor, a data calculation processor, a storage and display device, and a loop resistance measuring system, wherein the loop resistance measuring system is used to execute the loop resistance measuring method.

[0049] Compared with existing technologies, this invention has significant advantages: It acquires initial sample streams of voltage and excitation current through a three-channel synchronous acquisition architecture, forming a multivariate time series data base. Subsequently, a timing identifier is applied to the initial sample stream for inter-channel time alignment, eliminating time base errors and generating a synchronous voltage sample set with strictly aligned time axes. Then, through time-frequency domain feature extraction, peak values, root mean square values, and envelope mean values ​​are calculated to form an amplitude characterization subset. Combined with phase envelope tracking technology, an instantaneous phase sequence is derived, and a phase characterization subset is formed after consistency evaluation. Finally, a set of endpoint voltage amplitude and phase features comprehensively reflecting the relationship between signal energy and phase is constructed through noise-weighted fusion. In the core calculation stage, the instantaneous phase difference between channels is calculated based on the amplitude and phase features, and phase compensation parameters are generated to correct the voltage amplitude to eliminate system bias. The instantaneous loop resistance estimation sequence is obtained point-by-point using the corrected voltage difference and excitation current amplitude. This sequence undergoes steady-state estimation processing based on median filtering, robust statistical anomaly removal, and weighted averaging to suppress transient noise and outliers, outputting a reliable set of steady-state loop resistance estimates. Finally, at the diagnostic level, statistical modeling of the dataset is performed. By calculating the relative deviation sequence of resistance values ​​at each interface, the mean and standard deviation are extracted to form the imbalance rate data. Combined with confidence scores based on the consistency of measurements over multiple time periods, a composite judgment index is constructed to achieve automatic identification and prioritization of contact anomalies, outputting a structured anomaly list. This invention solves the problems of low measurement accuracy and insufficient anomaly identification capability caused by poor synchronization, phase errors, and lack of in-depth analysis in traditional methods by constructing a complete data chain covering high-precision synchronous sampling, adaptive signal processing, phase compensation calculation, and intelligent statistical diagnosis. This improves the accuracy, anti-interference ability, and intelligent level of condition diagnosis of loop resistance measurement.

[0050] The present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of a loop resistance measurement method in one embodiment.

[0052] Figure 2 This is a schematic diagram of the estimated interface resistance value in one embodiment.

[0053] Figure 3 This is a schematic diagram of a resistance measurement circuit in one embodiment. Detailed Implementation

[0054] The technical method of this invention 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 invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention.

[0055] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0056] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0057] In one embodiment, combined Figures 1 to 3 A method for measuring loop resistance is provided, the method comprising the following steps:

[0058] S1: Loop data is acquired at each interface of the disconnector using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream.

[0059] S2: Implement timing marking on the initial sampling stream to complete the identification of the sample start and end points, complete the time alignment between channels, and output a synchronous voltage sample set;

[0060] S3: Construct a set of endpoint voltage amplitude and phase characteristics from the synchronous voltage sample set;

[0061] S4: Based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling flow, the voltage difference after phase compensation is divided by the corresponding excitation current amplitude to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment, and the steady-state estimation is performed on the instantaneous estimation sequence to generate a set of estimated loop resistance values.

[0062] S5: Calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and judge contact abnormalities according to the preset judgment threshold and confidence rules. Output a list of contact abnormalities, including the corresponding interface identifier and measurement timestamp.

[0063] It is important to note that in S1, the differential isolation amplifier is used to suppress common-mode noise and provide electrical isolation, ensuring the accuracy and safety of voltage signal acquisition.

[0064] Meanwhile, in S4, phase compensation generates phase rotation parameters by calculating the instantaneous phase difference between the two channels to correct phase deviations caused by transmission delays or hardware differences, ensuring the accuracy of voltage difference calculation.

[0065] In this embodiment of the invention, reference Figure 1 The diagram shown is a flowchart illustrating the steps of a loop resistance measurement method according to the present invention. In this example, the loop resistance measurement method includes the following steps:

[0066] S1: Loop data is acquired at each interface of the disconnector using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream.

[0067] Preferably, in some embodiments, S1 includes the following steps:

[0068] At each interface of the disconnector switch using a three-channel wiring method, the excitation source outputs an excitation current signal at a preset frequency and amplitude, and injects it into each measurement branch through a signal distribution unit; under the action of the excitation signal, the differential isolation amplifier synchronously acquires the three-channel terminal voltage signal and the corresponding current response signal;

[0069] During the acquisition process, an in-phase reference compensation mechanism is adopted. The phase offset between the reference branch signal and the main measurement branch signal is compared and dynamically corrected to adjust the sampling amplitude-phase relationship. The synchronization control circuit times and numbers the sampling signal, and outputs the original voltage signal and initial sampling stream after synchronization calibration.

[0070] S2: Implement timing marking on the initial sampling stream to complete the identification of the sample start and end points, complete the time alignment between channels, and output a synchronous voltage sample set;

[0071] Preferably, in some embodiments, S2 includes the following steps:

[0072] Each sampling window is filtered using a programmable digital bandpass filter to remove the DC component;

[0073] Short-time energy spectra are calculated window by window after filtering, and stable periods are identified based on spectral stability criteria. The identified stable periods are arranged in time series to form a set of effective sampling segments.

[0074] The instantaneous voltage values ​​of each channel are extracted from the set of effective sampling segments in time alignment to form a candidate synchronization sample set, and the best one is selected to generate the synchronization voltage sample set.

[0075] In this embodiment of the invention, a time stamp is added to the original sample stream and it is framed at a fixed rate to clearly define the timestamp of each sample and divide the data into processable sampling windows. Each window is processed in the digital domain by a programmable bandpass filter to remove DC components to suppress low-frequency drift and DC bias, resulting in a bandwidth-controlled in-window time-domain signal. For each window, a short-time energy spectrum (e.g., a short-time estimate of the energy or power spectral density within the window) is calculated, and continuous peaks are identified using spectral stability criteria (e.g., variance / coefficient of variation of energy or spectral shape, fluctuation range of spectral peak position and width, or signal-to-noise ratio threshold). During stable periods, only stable segments that meet the minimum duration are considered valid sampling segments. Within these valid sampling segments, the instantaneous voltage values ​​of each channel are extracted according to the timestamp, and any time delay deviations are corrected by cross-correlation or time difference estimation between channels to generate a candidate synchronization sample set. The candidate set is then subjected to quality judgment (e.g., cross-correlation peak value between channels, consistency of amplitude of each channel, outlier removal) and is selected for retention according to priority (cross-correlation / signal-to-noise ratio / stable duration). Finally, a time-aligned synchronization voltage sample set that meets stability and consistency constraints is output.

[0076] In one implementation of the present invention, assuming a sampling rate of 10kHz per channel, the data is cut into a frame every 100ms (that is, 1000 samples per frame, with 50% overlap between frames, i.e., sliding once every 50ms).

[0077] First, each frame is filtered using a bandpass filter (e.g., retaining only the frequency band from 100Hz to 3kHz), and then a high-pass filter is used to remove DC drift below 10Hz. Next, an "energy value" is calculated for each frame (this is achieved by summing the squares of each sample within the frame, denoted as ). If the energy values ​​of three consecutive frames fluctuate very little—for example, the coefficient of variation of energy is less than 5% or the estimated signal-to-noise ratio is greater than 20dB—these three frames are linked together as a stable period; the stable period must last at least 200ms (that is, at least four frames) to be considered valid.

[0078] For these valid segments, instantaneous voltages are taken from each channel by timestamp and aligned using inter-channel cross-correlation: if the delay corresponding to the peak cross-correlation value is less than one sampling point (0.1ms), the alignment is considered complete; otherwise, the timestamps are fine-tuned using interpolation. The aligned candidate samples are then checked. If the cross-correlation coefficients between channels are all greater than 0.9 and the amplitude difference between each channel is within ±2%, this set of samples is added to the final "synchronization voltage sample set"; otherwise, the inconsistent samples are discarded.

[0079] Preferably, in some embodiments, calculating the short-time energy spectrum window by window after filtering and identifying stable periods based on spectral stability criteria includes:

[0080] For each window, a stability score is calculated based on the instantaneous signal-to-noise ratio index, where the stability score is obtained by combining the energy concentration and the variability rate of the spectrum shape over time within each window.

[0081] A threshold is set using stability scores, and windows that pass the stability test are selected to form a subset of stable windows.

[0082] A stable window subset is used as the source of the candidate synchronization sample set instead of the effective sampling segment set. The temporal distribution and spectral characteristics of the stable window subset constitute the effective sampling segment set.

[0083] In this embodiment of the invention, the short-time energy spectrum is first calculated on each bandpass filter window, and two types of metrics are extracted from the spectral domain: energy concentration (i.e., the ratio of energy falling in the target frequency band or several dominant frequency bands within the window to the total energy) and the rate of variation of the spectral shape over time (which can be represented by the statistical variation of indicators such as the spectral centroid, spectral peak position or spectral entropy of adjacent windows). The instantaneous signal-to-noise ratio of the signal within the window is estimated to reflect the measurement reliability.

[0084] The above metrics are normalized according to predetermined weights and synthesized into stability scores (range normalized to 0–1) to unify the comparison of spectral characteristics of different dimensions. Stability thresholds are set based on empirical or statistical methods (e.g., determining the optimal cutoff based on quantiles of historical data or ROC analysis), and windows with stability scores exceeding the threshold are selected to form a stable window subset. This subset is aggregated according to temporal continuity (merging adjacent windows or windows with gaps less than the threshold) to generate time periods that continuously meet the spectral stability conditions, and these time periods are used as the time source of effective sampling segments. Subsequently, instantaneous voltage samples of each channel are extracted from the stable window subset according to timestamps, and the micro-hour differences between channels are corrected by cross-correlation or time delay estimation to form a candidate synchronization sample set. Consistency checks are performed on the candidate set (e.g., cross-correlation peak values, amplitude consistency, and anomaly removal between channels) to output the final synchronization sample set obtained by spectral stability constraints for subsequent parameter estimation.

[0085] In one implementation of this invention, assuming a sampling rate of 10kHz per channel, the data is divided into frames every 100ms (1000 samples, with 50% overlap, sliding once every 50ms). For each frame, its energy percentage in the target frequency band (e.g., 100–3000Hz) is calculated. If this percentage is 85%, it indicates concentrated energy. Then, the change in the spectral centroid (frequency centroid) between the frame and the two frames before and after is observed. If the standard deviation of the centroids of the three frames is less than 50Hz, it indicates stable spectral shape.

[0086] Simultaneously estimate the instantaneous signal-to-noise ratio of this frame, for example, 22dB. Combine these three metrics with weights of 0.6, 0.3, and 0.1 to form a stability score of 0–1, which, in the example above, is approximately 0.83. If the threshold is set to 0.75, this frame is considered a "stable window". If at least four consecutive frames (i.e., at least 200ms) are considered stable, this period is considered a valid sampling segment.

[0087] Extract the instantaneous voltage of each channel from these stable segments according to the timestamp, and check their time alignment using cross-correlation: if the time delay corresponding to the peak of the cross-correlation is less than 0.1ms (one-tenth of a sampling point, or fine-tuned by interpolation), it is considered to be aligned; then check that the amplitude difference of each channel is within ±2% and the cross-correlation coefficient is greater than 0.9, and put this set of samples into the final synchronization voltage sample set; otherwise, discard it.

[0088] Preferably, the set of effective sampling segments also includes:

[0089] The adaptive window length parameter is automatically determined based on the instantaneous signal-to-noise ratio and spectral distribution of samples in the candidate synchronization sample set;

[0090] The adaptive window length parameter is used to adjust the temporal distribution of the sampling window, and the adaptive window length parameter is passed to the passband parameter of the bandpass filter to adjust the spectral characteristics of the filter bandwidth.

[0091] Under low instantaneous signal-to-noise ratio conditions, the sampling window is expanded according to the adaptive window length parameter to enhance statistical stability, while under high instantaneous signal-to-noise ratio conditions, the sampling window is shortened to improve temporal resolution.

[0092] The adaptive window length parameter is recorded as a set of valid sampling segments.

[0093] In embodiments of the present invention, the instantaneous signal-to-noise ratio (SNR) and spectral distribution characteristics (such as spectral concentration or spectral entropy) of each group of samples in the candidate synchronization sample set are used as inputs. A window length parameter (i.e., adaptive window length) is automatically determined and output through a preset mapping strategy or adaptive rule. The mapping can be based on experience-based table lookup, piecewise threshold mapping, or continuous monotonic function to map SNR and spectral distribution to the window length space.

[0094] The adaptive window length is then used for real-time adjustments in two aspects: First, it reconfigures the length and sliding step of the time-domain sampling window to match the window's temporal distribution with the current signal's statistical characteristics (e.g., extending the window in a low SNR scenario to increase the number of samples and improve statistical reliability, and shortening the window in a high SNR scenario to enhance time resolution); Second, it transmits the adaptive window length parameter as a control variable to the digital bandpass filter to adjust the passband width or filter order (e.g., widening the passband appropriately to accommodate more spectral energy when the window length is increased, and narrowing the passband to suppress out-of-band noise when the window length is shortened), thereby ensuring that the spectral characteristics of the filtered signal are consistent with the window length.

[0095] To avoid frequent jitter, the mapping process should include a smoothing or hysteresis mechanism (moving average or exponential smoothing) and minimum / maximum window length constraints. Finally, the determined adaptive window length and its trigger time, the corresponding SNR and spectral distribution statistics are recorded as metadata in the set of effective sampling segments.

[0096] In one implementation of the present invention, the system default window length is 100ms (1000 samples per frame, sampling rate 10kHz), and the defined window length can be adaptively adjusted between 50ms and 500ms;

[0097] When the instantaneous SNR of the candidate samples is below 10dB and the spectral concentration is below 70%, the mapping rule sets the window length to 400ms (to expand the window and obtain more samples for stable statistics); when the SNR is between 10 and 20dB and the spectral concentration is between 70% and 85%, the window length is set to 200ms; when the SNR is above 20dB and the spectral concentration is above 85%, the window length is shortened to 50–100ms to improve temporal resolution.

[0098] If the window length changes from 100ms to 400ms, the passband of the bandpass filter can be slightly widened from 100–3000Hz to 80–3200Hz to retain more spectral information; if the window length is shortened, the passband will be narrowed back to 100–3000Hz, and the filter order can be increased to enhance noise suppression. After each adjustment, records such as "window length = 200ms, trigger time = 12:34:56.789, SNR = 15dB, spectral concentration = 78%" are written to the metadata of the effective sampling segment set.

[0099] It is important to note that the choice of a 10kHz sampling rate is based on the Nyquist sampling theorem, ensuring coverage of the main frequency components of the excitation signal (typically below 5kHz) while avoiding aliasing. The 100ms window length takes into account signal stability and real-time requirements, striking a balance between statistical reliability and time resolution.

[0100] S3: Construct a set of endpoint voltage amplitude and phase characteristics from the synchronous voltage sample set;

[0101] Preferably, in some embodiments, S3 specifically includes:

[0102] For each effective sampling segment within the synchronous voltage sample set, the peak value, root mean square value, and envelope mean value are calculated by channel to form an amplitude characterization subset;

[0103] Phase envelope tracking is applied to each effective sampling segment to obtain an instantaneous phase sequence. Time-domain averaging and phase consistency evaluation are performed on the phase sequences of each channel within the same effective sampling segment to form a phase characterization subset.

[0104] The amplitude characterization subset and the phase characterization subset are merged and noise-weighted to generate the terminal voltage amplitude and phase feature set.

[0105] In embodiments of the present invention, time-domain amplitude statistics and envelope analysis are performed on each channel for each effective sampling segment to form an amplitude characterization subset—specifically including directly calculating the peak value (maximum absolute value of the sample), root mean square value (RMS, used to reflect the effective energy level), and envelope mean (obtained by analyzing the signal / Hilbert transform and then averaging it over time), and normalizing and robustly estimating these quantities (e.g., removing extrema or using median filtering to reduce the influence of single-point pulses); simultaneously, phase envelope tracking is performed on each channel for the same sampling segment to obtain the instantaneous phase sequence (by analyzing the signal phase, unwrapping the phase, and performing small... (Window smoothing), then statistically summarize the instantaneous phase of each channel in the time domain (e.g., calculate the phase mean, the mean square or circular square of the phase difference sequence, etc., to form a phase characterization subset); when merging amplitude and phase characterization, first assign a weight to each feature based on the instantaneous signal-to-noise ratio or in-band noise spectral density of the channel (the weight can be normalized by linearizing the SNR or defined by 1 / (noise variance)), and standardize each feature according to its scale before merging to avoid dimensional bias. Finally, the amplitude and phase features are weighted and fused according to the set noise weighting strategy and the amplitude and phase feature set of the terminal voltage is output.

[0106] In one implementation of the present invention, the sampling rate is 10kHz, an effective sampling segment is 200ms long (2000 samples), and the three-channel data are obtained respectively on this segment:

[0107] Channel 1 peak voltage is 3.2V, RMS voltage is 1.05V, and envelope average voltage is 1.10V.

[0108] Channel 2 peak voltage 2.9V, RMS voltage 0.98V, envelope mean voltage 1.00V;

[0109] Channel 3 peak voltage 3.0V, RMS voltage 1.02V, envelope mean voltage 1.03V.

[0110] After obtaining the analytic signal using Hilbert transform and calculating the instantaneous phase, the phase mean values ​​of the three channels within this segment are 12.4°, 13.1°, and 12.8°, respectively. The standard deviation of the phase difference sequence is calculated to be 0.9° (indicating high phase consistency). The instantaneous SNR within each channel segment is then measured: Channel 1 = 22dB, Channel 2 = 18dB, and Channel 3 = 20dB. These three SNRs are first linearized and normalized (e.g., weighted to 0.42, 0.30, and 0.33). Then, a weighted average is calculated for each amplitude and phase characteristic – for example, the weighted RMS ≈ 0.42·1.05 + 0.30·0.98 + 0.33·1.02 ≈ 1.02V, and the weighted phase mean ≈ 12.7°. The phase consistency (0.9°) and overall confidence level (given by the weighted SNR) are recorded.

[0111] If the mean envelope of a certain channel suddenly deviates from the other two channels by more than 5%, or the standard deviation of the phase difference sequence exceeds a preset threshold, then the contribution of that channel in that segment is reduced or eliminated, and the weighted result is recalculated. The numerical amplitude and phase feature set obtained by the above processing is as follows.

[0112] S4: Based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling flow, the voltage difference after phase compensation is divided by the corresponding excitation current amplitude to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment, and the steady-state estimation is performed on the instantaneous estimation sequence to generate a set of estimated loop resistance values.

[0113] Preferably, in some embodiments, S4 specifically includes:

[0114] S41: For each valid sampling segment in the synchronous voltage sample set, extract the endpoint voltage amplitude and phase feature subsets of the corresponding two channels according to the interface, wherein the endpoint voltage amplitude and phase feature subsets are obtained from the endpoint voltage amplitude and phase feature set;

[0115] S42: Calculate the instantaneous phase difference between the two channels within the same effective sampling segment based on the amplitude and phase feature subset of the endpoint voltage and generate phase compensation parameters therefrom, wherein the phase compensation parameters constitute the phase compensation data;

[0116] S43: Apply phase compensation data to perform phase correction on the terminal voltage amplitude, obtain the phase-compensated terminal voltage amplitude sequence, and record the phase compensation metadata used for correction.

[0117] S44: Calculate the voltage difference between the two sides after phase compensation according to the interface in each effective sampling segment, and obtain the estimated value of the instantaneous loop resistance of the interface in the effective sampling segment by dividing the voltage difference by the corresponding excitation current amplitude point by point in time.

[0118] S45: Perform steady-state estimation processing on the instantaneous loop resistance estimation value sequence to suppress transient noise and outliers. The steady-state estimation processing includes, but is not limited to, median filtering within the window, outlier removal based on robust statistics, and weighted average calculation. The processing result is the steady-state loop resistance estimation value of the effective sampling segment.

[0119] S46: Summarize the estimated steady-state loop resistance values ​​of each effective sampling segment according to the time series to form a set of estimated loop resistance values.

[0120] In the embodiments of the present invention, within each effective sampling segment, the amplitude and phase feature subsets of the corresponding two channels are extracted from the generated amplitude and phase feature set of the endpoint voltage according to the interface, so as to ensure that the subsequent calculation uses the amplitude and phase information of the interface pair as input; for each pair of channel features, the phase difference of the two channels at each moment is directly calculated based on the instantaneous phase sequence and a phase compensation parameter is generated therefrom (this parameter can be expressed as time delay / phase rotation amount or complex phase factor), and the phase compensation parameter is stored as phase compensation data in time sequence; then the phase compensation data is applied to the analytical signal or complex spectrum representation of the endpoint voltage, and the phases of the two channels are aligned by phase rotation / time delay correction, thereby obtaining the phase-compensated endpoint voltage amplitude sequence (in actual implementation, the phase correction is achieved by reconstructing the real part of the analytical signal or taking the real part after phase rotation), and the phase compensation metadata (compensation amount, trigger time, original phase difference, etc.) used for correction is recorded;

[0121] Within each valid sampling segment, the voltage difference between the two sides after phase compensation is calculated hourly on an interface-by-hour basis, and then divided point by point according to the excitation current amplitude at the corresponding moment (i.e., instantaneous loop resistance = voltage difference / excitation current) to obtain an instantaneous loop resistance estimation sequence. Steady-state estimation is performed on this instantaneous sequence to suppress transient noise and outliers. Steady-state estimation may include median filtering within the window, weighted averaging based on robust statistics (such as removing outliers exceeding k times the standard deviation or MAD-based outlier removal), etc., to obtain the steady-state loop resistance estimation value for each valid sampling segment. Finally, the steady-state estimation values ​​of each segment are summarized by time series and the corresponding metadata is retained.

[0122] In one implementation of the present invention, assuming an effective sampling segment of 200ms (2000 samples, sampling rate 10kHz), the measured values ​​of the two channels at a certain moment are: instantaneous amplitude of 3.20V and instantaneous phase of 12.4° for the left channel; and instantaneous amplitude of 2.95V and instantaneous phase of 13.6° for the right channel.

[0123] The phase difference between the two channels is φL-φR=-1.2°, and the generated phase compensation parameter is the rotation amount that moves the right channel phase closer to the left channel by 1.2°.

[0124] Applying this phase rotation to the right channel's analytical signal (equivalent to fine-tuning the real part by cos(1.2°)), the corrected amplitude of the right channel is approximately 2.94935V; the voltage difference between the two sides is approximately 0.25065V. If the excitation current amplitude at this moment is 5.00A, then the instantaneous loop resistance at this moment is estimated to be approximately 0.05013Ω.

[0125] After obtaining the instantaneous resistance sequence for 2000 time points in the entire segment, short-term pulse interference is first removed using a median filter, for example, at 51 points. Then, a weighted average is performed based on removing outlier samples that exceed 3 times the MAD (median absolute deviation). Assuming the median in the segment after filtering is 0.05010Ω, and the steady-state estimate after removing outliers and weighting is 0.05012Ω, metadata such as "start and end of time period, number of samples 2000, steady-state loop resistance = 0.05012Ω, confidence level given by the average SNR in the segment = 19dB, number of outlier samples removed = 7" is written into the record.

[0126] like Figure 2 The image shows the estimated resistance values ​​for the interfaces. A comparison of the measured resistance values ​​for interfaces A (0.05012Ω), B (0.06040Ω), and C (0.04995Ω) is also shown.

[0127] Preferably, in some embodiments, S4 further includes:

[0128] Two working modes, constant amplitude AC excitation and pulse amplitude excitation, are preset on the excitation source. The measurement system obtains the first loop resistance estimation set and the second loop resistance estimation set by parallel constant amplitude response analysis sub-process and pulse response analysis sub-process, respectively.

[0129] The difference significance extraction unit calculates the difference between the first loop resistance estimation set and the second loop resistance estimation set item by item, and identifies the set of interfaces with significant differences based on the statistical significance criterion.

[0130] The discrepancy results are incorporated as metadata into the set of loop resistance estimates.

[0131] It is important to note that the constant amplitude AC excitation mode is suitable for steady-state resistance measurement, providing continuous and stable excitation; while the pulse amplitude excitation mode is used to capture transient responses and identify anomalies such as loose contacts or oxidation. The system enhances the measurement method by processing the responses of both modes in parallel.

[0132] S5: Calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and judge contact abnormalities according to the preset judgment threshold and confidence rules. Output a list of contact abnormalities, including the corresponding interface identifier and measurement timestamp.

[0133] Preferably, in some embodiments, step S5, which involves calculating the imbalance rate of each interface based on the set of estimated loop resistance values ​​and determining contact abnormalities according to a preset judgment threshold and confidence rules, includes:

[0134] For each interface, a relative deviation sequence is constructed based on the set of estimated loop resistance values, where the relative deviation sequence is the relative difference between the estimated loop resistance value of the interface and the mean of the set of estimated loop resistance values ​​of the interface;

[0135] Calculate the mean and standard deviation of the deviation sequence, and use the joint criterion of the mean and standard deviation to form the imbalance rate data;

[0136] The imbalance rate data is compared with a preset judgment threshold and combined with the confidence score to generate contact anomaly records, thereby constructing a contact anomaly list, which includes the corresponding interface identifier and measurement timestamp.

[0137] In embodiments of the present invention, for each interface, the time-series estimated value of the interface within the set of loop resistance estimates is first constructed into a relative deviation sequence—that is, the difference between the estimated value at each moment and the mean of the entire segment of the interface is divided by the mean to obtain a dimensionless relative deviation sequence; then, statistical characteristics (including at least the mean and standard deviation, and if necessary, the median and MAD are also calculated for robustness testing) are calculated for the deviation sequence, and unbalance rate data is constructed using the joint judgment criteria of the mean and standard deviation. For example, the unbalance rate is defined as the relative value of the standard deviation of the deviation sequence (or a weighted combination of the mean and standard deviation), and the sample size, time distribution, and average SNR within the segment are recorded as confidence inputs; the unbalance rate is compared with a pre-set judgment threshold, and a contact anomaly judgment is generated and written into an anomaly record according to the confidence rules (e.g., when the sample size is insufficient or the average SNR is low, the confidence is reduced or the threshold is increased; when the SNR is high and the sample size is large, the original threshold is used for judgment). The anomaly record includes the interface identifier, judgment timestamp, unbalance rate value, mean and standard deviation, sample size, confidence score, and threshold used.

[0138] In one implementation of the present invention, it is assumed that the estimated set of loop resistance values ​​for a certain interface consists of five numbers (unit: Ω):

[0139] 0.05010, 0.05012, 0.05020, 0.05100, 0.04980.

[0140] Find the mean: add the five numbers together to get 0.050244Ω.

[0141] The relative skewness of each item is calculated as (item - mean) ÷ mean, and this is done item by item.

[0142] The first deviation is -0.2866%.

[0143] The second deviation was -0.2468%.

[0144] The third deviation was -0.0876%.

[0145] The fourth deviation is +1.5047%;

[0146] The fifth deviation is -0.8837%. The third step is to calculate the statistic of the deviation series: using the sample standard deviation (which is more sensitive to sample fluctuations), we get approximately 0.000449Ω, which is converted to a relative standard deviation of ≈0.8938%.

[0147] If the preset imbalance rate threshold is 1.0% (i.e., a relative standard deviation > 1.0% is considered abnormal) and the confidence rule requires an average SNR ≥ 15dB and a sample size ≥ 5 for strict judgment to be performed; in this example, the relative standard deviation is approximately 0.8938%, which is less than 1.0%, and assuming an average SNR of 19dB and a sample size of 5, then according to the rule, this interface will not be marked as a contact anomaly; if another interface calculates a relative standard deviation of 1.6% (exceeding the threshold) and its average SNR is also higher than the threshold, then a contact anomaly record will be generated, with the record content as an example:

[0148] {Interface ID:X, Timestamp:2025-10-31T12:34:56Z, Imbalance Rate:1.60%, Mean:0.05024Ω, Standard Deviation:0.00080Ω, Sample Size:8, Average SNR:22dB, Decision Threshold:1.0%, Confidence:High}.

[0149] Preferably, in some embodiments, the contact anomaly list specifically includes:

[0150] Calculate the confidence score based on the consistency of loop resistance estimates across multiple independent valid sampling segments using the same interface;

[0151] A composite judgment index is formed by combining imbalance rate data and confidence score, and the items in the contact anomaly list are assigned priority ranking based on the composite judgment index.

[0152] Priority sorting is used by portable terminals or remote monitoring terminals to output measurement reports in order of priority. The measurement reports include the corresponding interface identifier and measurement timestamp.

[0153] In embodiments of the present invention, a consistency analysis is performed on the steady-state loop resistance estimation sequence obtained from multiple independent valid sampling segments of the same interface to generate a confidence score. Specifically, the estimation values ​​of each segment are summarized over time, robust statistics (such as median, MAD, or mean and sample standard deviation) are calculated, and the sample size, time-series distribution density, and average SNR within the segment are used as inputs. Through mapping rules (table lookup, segment threshold, or continuous mapping function), the consistency index (such as the inverse ratio of relative standard deviation or coefficient of variation) and measurement reliability are integrated into a normalized confidence score (0–1). Subsequently, the imbalance rate data (which can be relative standard deviation, MAD, or mean and sample standard deviation) are analyzed. The standard deviation (or scaled imbalance index) and confidence score are combined according to a preset weighting strategy to form a composite judgment index (e.g., composite score = w1·standardized imbalance rate + w2·(1−confidence level), or other monotonic mapping to ensure that the score is consistent with the severity of the anomaly). Based on the composite judgment index, all anomaly candidates are sorted in descending order to form a priority list. During the sorting process, the original statistics and metadata used for judgment (sample size, time window, average SNR, mean / variance of each segment, etc.) should be recorded. For cases with low confidence or insufficient sample size, a conservative strategy should be applied (e.g., increasing the trigger threshold or marking it as "retest required").

[0154] Finally, the contact anomaly list, sorted by priority, will be used to generate measurement reports for portable terminals or remote monitoring terminals. The report entries include interface identifier, priority level, judgment timestamp, imbalance rate, uncertainty / confidence score, relevant statistics, and recommended review actions.

[0155] In one implementation of the present invention, the steady-state resistance values ​​obtained from multiple sampling segments for each of the three interfaces (A, B, and C) are as follows:

[0156] A:[0.05010,0.05012,0.05008,0.05015]Ω;

[0157] B: [0.0605, 0.0610, 0.0598]Ω;

[0158] C:[0.0499,0.0500]Ω.

[0159] First, calculate the mean and relative standard deviation (or coefficient of variation) for each interface, and combine this with the sample size and average SNR to obtain the confidence score (for example, using empirical mapping: A's coefficient of variation ≈ 0.13%, sample size 4, average SNR = 20dB → confidence score 0.94; B's coefficient of variation ≈ 1.03%, sample size 3, average SNR = 18dB → confidence score 0.78; C's coefficient of variation ≈ 0.14%, sample size 2, average SNR = 12dB → confidence score 0.60; due to the small sample size and low SNR, the confidence score is reduced). Then, take the imbalance rate (for example, using relative standard deviation or pre-scaling to the 0–1 interval): A = 0.13% → 0.13 (standardized 0.13), B = 1.03% → 1.03, C = 0.14% → 0.14. Set the combination rule as composite score = 0.7 × (standardized imbalance rate) + 0.3 × (1 − confidence score), then we get:

[0160] A composite score ≈ 0.109;

[0161] B composite fraction 0.787;

[0162] C composite score ≈ 0.218.

[0163] The composite scores are sorted in descending order to obtain priorities B (high), C (medium), and A (low). Example of a measurement report (output by priority): The first report is for interface B.

[0164] {Interface ID: B, Timestamp: 2025-10-31T12:40:02Z, Priority: High, Imbalance Rate: 1.03%, Confidence: 0.78, Sample Size: 3, Remarks: On-site verification recommended};

[0165] The second entry is C: {Interface ID: C, Timestamp: 2025-10-31T12:39:50Z, Priority: Medium, Imbalance Rate: 0.14%, Confidence: 0.60, Sample Size: 2, Remarks: Sample Size is Too Small};

[0166] The third entry is A: {Interface ID: A, Timestamp: 2025-10-31T12:41:10Z, Priority: Low, Imbalance Rate: 0.13%, Confidence: 0.94, Number of Samples: 4}.

[0167] See Figure 3 A loop resistance measuring device is provided, including an excitation source, a voltage acquisition circuit, a synchronization control circuit, a signal processor, a data processing processor, a storage and display device, and a loop resistance measuring system.

[0168] The excitation source applies an AC excitation signal with known amplitude and phase characteristics to the circuit under test. Its output is input to the circuit under test via an electrical interface, forming a stable excitation current. The voltage acquisition circuit synchronously acquires the terminal voltage signals on both sides of each interface in the circuit under test, and performs high-precision amplification, filtering, and analog-to-digital conversion on the voltage signals to generate a continuous voltage sampling sequence. The synchronization control circuit coordinates the timing of the excitation source and the voltage acquisition circuit, ensuring that current sampling and voltage sampling are strictly aligned in time, thereby maintaining amplitude and phase consistency in subsequent calculations.

[0169] The acquired voltage and current sample data are input to a signal processor, which performs noise suppression, baseline drift correction, phase correction, and amplitude extraction to obtain a high-quality amplitude and phase feature dataset. Subsequently, this processed data is transmitted to a data processing processor, which extracts the phase difference of the endpoint voltages, calculates instantaneous resistance values, performs steady-state estimation, and generates a set of estimated loop resistance values ​​according to predefined calculation logic. The data processing processor is also responsible for establishing imbalance rate indices and confidence level judgments between the interfaces at the statistical level to output a structured loop resistance measurement data table.

[0170] The storage and display device receives and saves the output results of the data calculation processor, presenting the estimated loop resistance values, unbalance rate data, anomaly indicators, and measurement timestamps in a visual manner. It can also upload these results to a host computer or remote monitoring system via an interface for further analysis and archiving. The entire measurement system forms a closed-loop structure from excitation signal generation, synchronous acquisition, signal processing, data calculation to result output, with each unit interconnected via a digital bus and timing synchronization signals.

[0171] Meanwhile, it should be noted that the excitation source is used to generate AC or pulse excitation signals with preset frequency and amplitude to ensure stable excitation current; the voltage acquisition circuit includes a differential isolation amplifier and an analog-to-digital converter to achieve high-precision voltage sampling; the synchronization control circuit coordinates the synchronization of excitation and acquisition through timing markers and reference signals; the signal processor is responsible for filtering, phase correction, and feature extraction; the data processing processor executes resistance calculation, steady-state estimation, and anomaly detection algorithms; and the storage and display device is used for data storage, visualization, and remote transmission.

[0172] Among them, the process starting point, the excitation source -- 100 generates an excitation signal.

[0173] Data acquisition and preprocessing: The three-channel acquisition module--110 acquires the raw data, and the signal synchronization module (120) completes the alignment and synchronization.

[0174] Feature extraction: The endpoint voltage feature extraction module--130 extracts key amplitude and phase features from the synchronization data.

[0175] The core calculation involves extracting features that are then sent to the resistance calculation module -- 140, which calculates the final loop resistance value.

[0176] Intelligent diagnosis: The resistance calculation results are sent to the anomaly detection module--200 to automatically complete the status diagnosis and anomaly identification.

[0177] In summary, this invention solves the problems of low measurement accuracy and insufficient anomaly identification capability caused by poor synchronization, phase error and lack of in-depth analysis in traditional methods by constructing a complete data chain covering high-precision synchronous sampling, adaptive signal processing, phase compensation calculation and intelligent statistical diagnosis. It improves the accuracy, anti-interference ability and intelligent level of loop resistance measurement.

[0178] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0179] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. A method for measuring loop resistance, characterized in that, The method includes the following steps: Step 1: Loop data is acquired at each interface of the disconnector using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream. Step 2: Implement timing identification on the initial sampling stream to complete the identification of the sample start and end points, complete the time alignment between channels, and output a synchronous voltage sample set; Step 3: Construct a set of endpoint voltage amplitude and phase features from the set of synchronous voltage samples; Step 4: Based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling current, the voltage difference after phase compensation is divided by the corresponding excitation current amplitude to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment, and the steady-state estimation is performed on the instantaneous estimation sequence to generate a set of loop resistance estimated values. Step 5: Calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and judge contact abnormalities according to the preset judgment threshold and confidence rules. Output a list of contact abnormalities, including the corresponding interface identifier and measurement timestamp.

2. The loop resistance measurement method according to claim 1, characterized in that, Step 2 specifically includes: Step 2-1: Filter each sampling window using a programmable digital bandpass filter to remove the DC component; Step 2-2: Calculate the short-time energy spectrum window by window after filtering and identify stable periods based on the spectrum stability criterion. Then, construct a set of effective sampling segments by time series for the identified stable periods. Steps 2-3: Extract the instantaneous voltage values ​​of each channel from the set of effective sampling segments according to time alignment to form a candidate synchronization sample set, and select the best to generate the synchronization voltage sample set.

3. The loop resistance measurement method according to claim 2, characterized in that, Step 2-2, which involves calculating the short-time energy spectrum window by window after filtering and identifying stable periods based on the spectral stability criterion, specifically includes: For each sampling window, a stability score is calculated based on the instantaneous signal-to-noise ratio; the stability score is obtained by combining the energy concentration and the variability rate of the spectral shape over time within each sampling window. A threshold is set using stability scores, and windows that pass the stability test are selected to form a subset of stable windows; A stable window subset is used as the source of the candidate synchronization sample set instead of the effective sampling segment set. The temporal distribution and spectral characteristics of the stable window subset constitute the effective sampling segment set.

4. The loop resistance measurement method according to claim 3, characterized in that, The set of valid sampling segments also includes: The adaptive window length parameter is automatically determined based on the instantaneous signal-to-noise ratio and spectral distribution of samples in the candidate synchronization sample set; The adaptive window length parameter is used to adjust the temporal distribution of the sampling window, and the adaptive window length parameter is passed to the passband parameter of the bandpass filter to adjust the spectral characteristics of the filter bandwidth. Under low instantaneous signal-to-noise ratio conditions, the sampling window is expanded according to the adaptive window length parameter to enhance statistical stability, while under high instantaneous signal-to-noise ratio conditions, the sampling window is shortened to improve temporal resolution. The adaptive window length parameter is recorded in the set of valid sampling segments.

5. The loop resistance measurement method according to claim 1, characterized in that, Step 3 specifically includes: Step 3-1: For each valid sampling segment within the synchronous voltage sample set, calculate the peak value, root mean square value, and envelope mean value by channel to form an amplitude characterization subset; Step 3-2: Phase envelope tracking is performed on each effective sampling segment to obtain the instantaneous phase sequence. Time-domain averaging and phase consistency evaluation are performed on the phase sequences of each channel within the same effective sampling segment to form a phase characterization subset. Step 3-3: Merge the amplitude characterization subset and the phase characterization subset, and perform noise weighting to generate the terminal voltage amplitude and phase feature set.

6. The loop resistance measurement method according to claim 1, characterized in that, Step 4 specifically includes: Step 4-1: For each valid sampling segment in the synchronous voltage sample set, extract the endpoint voltage amplitude and phase feature subsets of the corresponding two channels according to the interface, wherein the endpoint voltage amplitude and phase feature subsets are obtained from the endpoint voltage amplitude and phase feature set. Step 4-2: Calculate the instantaneous phase difference between the two channels within the same effective sampling segment based on the amplitude and phase feature subset of the endpoint voltage, and generate phase compensation parameters therefrom, wherein the phase compensation parameters constitute the phase compensation data; Step 4-3: Apply phase compensation data to perform phase correction on the terminal voltage amplitude, obtain the phase-compensated terminal voltage amplitude sequence, and record the phase compensation metadata used for correction. Step 4-4: Calculate the voltage difference between the two sides after phase compensation according to the interface in each effective sampling segment, and obtain the estimated value of the instantaneous loop resistance of the interface in the effective sampling segment by dividing the voltage difference by the corresponding excitation current amplitude point by point in time, based on the corresponding initial sampling current amplitude in the effective sampling segment. Steps 4-5 involve performing steady-state estimation processing on the instantaneous loop resistance estimation sequence to suppress transient noise and outliers. The steady-state estimation processing includes median filtering within the window, outlier removal based on robust statistics, and weighted average calculation. The processing result is the steady-state loop resistance estimation value of the effective sampling segment. Steps 4-6: Summarize the estimated steady-state loop resistance values ​​of each effective sampling segment according to the time series to form a set of estimated loop resistance values.

7. The loop resistance measurement method according to claim 1, characterized in that, Step 5 calculates the imbalance rate of each interface based on the estimated loop resistance values ​​and determines contact anomalies according to preset judgment thresholds and confidence rules, specifically including: For each interface, a relative deviation sequence is constructed based on the set of estimated loop resistance values, where the relative deviation sequence is the relative difference between the estimated loop resistance value of the interface and the mean of the set of estimated loop resistance values ​​of the interface; Calculate the mean and standard deviation of the biased series, and form the imbalance rate data using the joint judgment criterion of the mean and standard deviation; The imbalance rate data is compared with a preset judgment threshold and combined with the confidence score to generate contact anomaly records, thereby constructing a contact anomaly list, which includes the corresponding interface identifier and measurement timestamp.

8. The loop resistance measurement method according to claim 7, characterized in that, The specific details of the access exception list in step 5 are as follows: Calculate the confidence score based on the consistency of loop resistance estimates across multiple independent valid sampling segments using the same interface; A composite judgment index is formed by combining imbalance rate data and confidence score, and the items in the contact anomaly list are assigned priority ranking based on the composite judgment index. Priority sorting is used to output priority-based measurement reports from portable terminals or remote monitoring terminals. The measurement reports include the corresponding interface identifier and measurement timestamp.

9. A loop resistance measurement system, characterized in that, For performing the loop resistance measurement method as described in any one of claims 1 to 8, the loop resistance measurement system comprises: The three-channel acquisition module is used to acquire loop data at each interface of the disconnector switch using a three-channel wiring method. Excitation current is injected by the excitation source and the original three-channel voltage signal and initial sampling stream are acquired by the differential isolation amplifier to obtain the initial sampling stream. The signal synchronization module is used to implement timing identification on the initial sampling stream to complete the identification of the start and end points of the samples, and to complete the time alignment between channels and output a set of synchronized voltage samples. The endpoint voltage feature extraction module is used to construct an endpoint voltage amplitude and phase feature set from the synchronous voltage sample set; The resistance calculation module is used to calculate the instantaneous estimated value of the loop resistance of each interface in each effective sampling segment based on the set of amplitude and phase characteristics of the endpoint voltage and the initial sampling current, by dividing the voltage difference after phase compensation by the corresponding excitation current amplitude, and to perform steady-state estimation on the instantaneous estimation sequence to generate a set of loop resistance estimated values. The anomaly detection module is used to calculate the unbalance rate of each interface based on the set of estimated loop resistance values, and to judge contact anomalies according to preset judgment thresholds and confidence rules. It outputs a list of contact anomalies, including the corresponding interface identifier and measurement timestamp.

10. A loop resistance measuring device, characterized in that, It includes an excitation source, a voltage acquisition circuit, a synchronization control circuit, a signal processor, a data processing processor, a storage and display device, and a loop resistance measurement system, wherein the loop resistance measurement system is used to perform the loop resistance measurement method as described in any one of claims 1 to 8.

Images