Detection method and detection system for digital output synchronism of mutual inductor

Abstract

The invention discloses a mutual inductor digital output synchronism detection method and system, particularly relates to the field of electric power engineering detection, and is used for solving the problem of synchronism deviation concealment caused by phase-locked loop drift of digital output of an electronic mutual inductor in an intelligent substation. A uniform time frame is utilized to penetrate through the whole process, synchronization deviation is continuously mapped into a phase migration track, long-term drift and instantaneous jitter are analyzed on two scales by means of phase shift stacking quantity and pulse dispersion rate, two kinds of heterogeneous features are converted into a single drift tradeoff coefficient through Gaussian process regression, and a stability domain threshold value is dynamically revised according to the single drift tradeoff coefficient; after the threshold is triggered, the minimum adjustment amount is immediately issued to the phase-locked loop and is synchronously verified, so that the clock adjustment action and the drift monitoring are mutually nested; the protection device continuously receives the stable sampling sequence, the tripping misjudgment rate is remarkably converged, the inspection period is prolonged, and additional hardware investment is not needed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power engineering detection, more particularly, the present application relates to a kind of mutual inductor digital output synchronism detection method and detection system. BACKGROUND

[0002] In modern intelligent substation, electronic mutual inductor continuously sends sampling value message by optical fiber, and protection device determines fault type and action sequence according to the message. The internal mutual inductor maintains time base by phase-locked loop, and encapsulates and outputs cache data into frame within a set interval. Field inspection usually uses closed-loop calibration device to connect mutual inductor output and reference channel into the same second pulse. Unified time scale can align two data instantaneously, and the calibration result only shows average delay, so the slight swing in message interval is completely hidden.

[0003] The mutual inductor clock runs independently for a long time, and the phase-locked loop continuously accumulates slight drift due to temperature drift and aging, and the sampling cache queue also injects random delay in the message assembly stage. After the superposition of the two factors, the actual sending interval of the message appears irregular dispersion; the closed-loop calibration automatically eliminates such dispersion by sharing the second pulse, and the detection process always shows ideal synchronization. After running for a period of time, the interval jitter continuously amplifies, the direction-finding threshold of the protection device is quietly pushed away from the safety zone, and finally induces mis-trip, while the inspection record lacks early warning. The present application aims to directly reveal the real jitter characteristics of the digital output of the mutual inductor by asynchronous capture and statistical analysis, and evaluate the synchronization reliability from the source. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the present application provides a kind of mutual inductor digital output synchronism detection method and detection system, to anchor data acquisition starting point with independent external time base, utilize unified time frame throughout the whole process, first map the synchronization deviation into phase walk trajectory continuously, then analyze long-term drift and instantaneous jitter in double scale by using the amount of phase shift superposition and pulse dispersion rate, and convert the two types of heterogeneous characteristics into a single drift trade-off coefficient by Gaussian process regression, and dynamically revise the threshold value of the stable domain; the threshold value triggers the minimum adjustment amount to the phase-locked loop and synchronously verifies it immediately, so that the clock adjustment action and the drift monitoring are nested with each other, to solve the problems raised in the above background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical scheme: A kind of mutual inductor digital output synchronism detection method, comprising the steps of: S1: synchronously intercepts mutual inductor digital message and reference pulse under the constraint of independent external time base, and records to unified time frame; S2: get synchronization deviation sequence by using unified time frame; S3: Map the synchronization deviation sequence to a phase walk trajectory, and construct the phase shift stacking amount by accumulating the offset through a sliding window. Combine this with the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generate the drift tradeoff coefficient through Gaussian process regression. S4: Inject the drift tradeoff coefficient into the synchronization stability region to form a deviation weight vector, and dynamically adjust the threshold accordingly; S5: When the deviation weight vector reaches the threshold, the phase-locked loop adjustment command is triggered, and the process returns to step S1 to complete the acquisition verification and threshold iteration.

[0006] In a preferred embodiment, step S1 includes deploying an independent external time base as a data acquisition constraint source, connecting the acquisition device between the current transformer output port and the reference pulse source, generating a trigger signal from the independent external time base to initiate the current transformer digital message sequence and reference pulse sequence interception operation, buffering the current frame and timestamp and the current pulse and timestamp at the beginning of each cycle to ensure that the interception window covers the complete message interval cycle, recording the sequence to a unified time frame, and using the independent external time base cycle as the unit of a two-dimensional structure, the first dimension stores the current transformer digital message sequence frame content and timestamp pair, and the second dimension stores the reference pulse sequence pulse content and timestamp pair, merging the two sequences by aligning the timestamps.

[0007] In a preferred embodiment, step S2 includes extracting the transformer timestamp sequence and the reference timestamp sequence from the unified time frame, calculating the difference between the elements of the transformer timestamp sequence and the elements of the reference timestamp sequence to form a synchronization deviation sequence, initializing the distribution container to accumulate and count the synchronization deviation sequence, updating the frequency of each deviation interval in the distribution container when a new set of unified time frame data is received, using a rolling update mechanism to retain the synchronization deviation sequence elements within the most recent fixed window, and defining the statistical distribution using the deviation value to frequency mapping function.

[0008] In a preferred embodiment, step S3 includes projecting a phase walk trajectory from the synchronization deviation sequence, converting the elements of the synchronization deviation sequence into phase values ​​using statistical distribution weights, forming a phase walk trajectory sequence by taking the cumulative deviation modulo 2π, applying a sliding window to the phase walk trajectory to calculate the absolute difference between adjacent elements and accumulating them to construct a phase shift stacking sequence, calculating the adjacent differences from the transformer timestamp sequence to form a message interval sequence, performing Mandel Segmentation on the message interval sequence to divide it into self-similar sub-segments, calculating the Hearst exponent of each sub-segment and taking the geometric mean as the pulse scattering rate.

[0009] In a preferred embodiment, the phase shift stacking sequence and pulse scattering rate are normalized and then input into a Gaussian process regression model, and drift tradeoff coefficients are generated by fitting a radial basis function kernel.

[0010] In a preferred embodiment, step S4 includes defining the synchronization stability region as a vector space corresponding to the number of statistical distribution intervals of the synchronization deviation sequence, and applying the drift tradeoff coefficient as a power exponent to the initial value of each dimension to form a deviation weight vector.

[0011] In a preferred embodiment, the deviation weight vector is calculated by applying the Tsallis entropy form to the normalized elements, summing the preset non-additive parameters, and then transforming the result to obtain the adjustment threshold.

[0012] In a preferred embodiment, step S5 includes real-time comparison of the deviation weight vector norm and the dynamic adjustment threshold, determining the impact by quantifying the deviation weight vector amplitude through the Holder norm, generating a minimum adjustment amount based on the maximum element of the deviation weight vector as a phase-locked loop adjustment command and sending it to the internal phase-locked loop of the transformer, repeatedly executing step S1 to obtain a new unified time frame, applying steps S2 to S4 to verify the adjustment effect on the new synchronization deviation sequence, and guiding iterative dynamic adjustment of the threshold based on the sign function of the difference obtained by subtracting the average value of the old synchronization deviation sequence from the average value of the new synchronization deviation sequence.

[0013] A digital output synchronization detection system for a current transformer includes: a data acquisition module, a deviation maintenance module, a drift fusion module, a threshold dynamic module, and an adjustment cycle module; The data capture module synchronously captures the digital messages and reference pulses of the current transformer under independent external time base constraints and records them to a unified time frame; The deviation maintenance module uses a unified time frame to obtain a synchronization deviation sequence; The drift fusion module maps the synchronization deviation sequence to a phase walk trajectory, and constructs the phase shift stacking amount by accumulating the offset through a sliding window. It combines the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generates the drift tradeoff coefficient through Gaussian process regression. The threshold dynamic module injects the drift tradeoff coefficient into the synchronization stability region to form a deviation weight vector, and dynamically adjusts the threshold accordingly. The adjustment loop module is used to trigger the phase-locked loop adjustment command when the deviation weight vector reaches the threshold, and then return to the data interception module to complete the acquisition verification and threshold iteration.

[0014] The technical effects and advantages of the present invention, a method and system for detecting the synchronization of digital output of a current transformer, are as follows: This invention uses an independent external time base to anchor the data acquisition starting point and a unified time frame throughout the entire process. First, synchronization deviation is continuously mapped into a phase walk trajectory. Then, phase shift stacking and pulse scattering rate are extracted from the dual scales to characterize long-term drift and instantaneous jitter. These two are then fused into a single drift tradeoff coefficient using Gaussian process regression. Based on this coefficient, the stability domain threshold is adjusted. Once the threshold is activated, a minimum range step command is immediately sent to the phase-locked loop (PLL) while simultaneously performing verification. Subsequently, clock adjustment behavior and drift detection behavior form a closed-loop cycle. Finally, a parallel clock adjustment behavior mechanism is constructed through a full-chain approach of acquisition, decomposition, prediction, and intervention. This achieves the effect of maintaining intra-frame compactness of the transformer's digital output in both low-load and high-load scenarios, and enables real-time suppression of the hidden erosion caused by external temperature drift and network congestion. This allows the protection device to continuously receive a stable sampling sequence, thereby reducing tripping errors, extending the inspection cycle, and requiring no additional hardware investment. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for detecting the synchronization of digital outputs of a current transformer according to the present invention. Figure 2 This is a schematic diagram of the structure of a current transformer digital output synchronization detection system according to the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Example 1: Figure 1 This invention provides a method for detecting the synchronization of digital outputs of a current transformer, comprising: S1: Under independent external time base constraints, synchronously intercept the digital messages and reference pulses of the current transformer and record them to a unified time frame.

[0018] S2: Obtain the synchronization deviation sequence using a unified time frame.

[0019] S3: Map the synchronization deviation sequence to a phase walk trajectory, and construct the phase shift stacking amount by accumulating the offset through a sliding window. Combine this with the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generate the drift tradeoff coefficient through Gaussian process regression.

[0020] S4: Inject the drift tradeoff coefficient into the synchronization stability domain to form a deviation weight vector, and dynamically adjust the threshold accordingly.

[0021] S5: When the deviation weight vector reaches the threshold, the phase-locked loop adjustment command is triggered, and the process returns to step S1 to complete the acquisition verification and threshold iteration.

[0022] Electronic instrument transformers continuously send sampled value messages to the protection devices of modern intelligent substations. Through this information, the protection devices can determine the fault type and action sequence. However, due to temperature drift and aging issues in the phase-locked loop inside the instrument transformer, the interval between sending sampled value messages will exhibit random and discrete phenomena. Under such circumstances, the traditional closed-loop testing method cannot truly verify the synchronization of the instrument transformer from the source.

[0023] To address the issues of implicit drift and sudden random delay that occur during long-term use of digital transformer outputs, a separate independent external time base locking method is used to replace the rejection effect caused by traditional automated verification, directly capturing the timing relationship between the digital message sequence and the standard pulse sequence within the actual measurement period.

[0024] Step S1 is divided into the following sub-steps: S1.1: Deploy an independent external time base as the acquisition constraint source. This independent external time base originates from a high-precision atomic clock or satellite timing module. The frequency stability of this independent external time base is superior to that of the acquisition device connected between the transformer output port and the reference pulse source within the transformer. The independent external time base generates a trigger signal, thereby initiating the interception operation of the transformer digital message sequence and the reference pulse sequence. The transformer digital message sequence refers to the continuously output sampled value frames within the fiber optic link of the electronic transformer; the reference pulse sequence refers to a standard second pulse or a sub-second level reference signal. This method uses a more accurate external reference to replace the transformer's internal clock, eliminating the problem of accumulated errors caused by the internal clock during acquisition. This ensures an absolute acquisition time base and prevents subsequent acquisition processes from being affected by changes in the transformer's internal clock, thus ensuring that all components of the time series are within a unified time frame.

[0025] S1.2: Based on the trigger signal of an independent external time base, the synchronous acquisition of the transformer digital message sequence and the reference pulse sequence is completed. At the start of each cycle of the independent external time base, the acquisition device buffers the current frame and timestamp of the transformer digital message sequence, as well as the current pulse and timestamp of the reference pulse sequence, ensuring that the width of the acquisition window can cover at least one complete message interval cycle; and the timestamp used adopts the absolute timing scale of the independent external time base. Using the trigger signal of the independent external time base, the original timing sequences of the two sequences are acquired at the same starting point to detect the true jitter of the transformer digital message sequence, eliminate the bias masking caused by automatic alignment, and ensure that after the acquired results are entered into the same time frame, each sequence obtained is an unbiased original dataset.

[0026] S1.3: Record the extracted transformer digital message sequence and reference pulse sequence into the same time frame. The time frame is a two-dimensional array consisting of an independent external time base period as one dimension, the frame content and corresponding timestamp of the transformer digital message in each period as another dimension, and the pulse content and timestamp of the reference pulse sequence as yet another dimension. After aligning the timestamps, the transformer digital message sequence elements and reference pulse sequence elements occupying the same position in the unified time frame share the same independent external time base marker. The above recording process places two different sequences into the same time coordinate system by merging them, which facilitates the direct calculation of the synchronization deviation sequence based on the time sequence and avoids gaps in analysis caused by different data islands. It is a data container containing all information. In the subsequent step S2, the synchronization deviation sequence can be extracted using this unified time frame to complete the stage determination of the source packet of the transformer digital output synchronization in the smart substation.

[0027] The sub-step described in this paper completes the unified sequence filling method for the original time sequence of digital message sequence and reference pulse sequence when the time frame is unified, and lays the foundation data for the unmasked synchronization deviation of subsequent steps. Compared with the implicit elimination of the interval jitter between messages by the traditional closed-loop verification, it demonstrates the true synchronization of the digital output of the instrument transformer in the smart substation.

[0028] Step S2 uses a unified time frame to calculate the synchronization deviation sequence and maintains a statistical distribution to quantify the timing difference between the message and the reference pulse, preparing data for subsequent drift analysis. This is because the unified time frame already captures the original time sequence, allowing for the direct acquisition of the most accurate jitter based on the deviation sequence measured in a unified manner. Real-time maintenance of the statistical distribution monitors deviation changes constantly, avoiding the neglect of dynamic changes caused by using static averages. For substation inspections, this allows for the early detection of potential synchronization faults.

[0029] The time-series mapping relationship between the digital message sequence number of the instrument transformer and the reference pulse sequence number is recorded in the unified time frame established in step S1. Based on this, the synchronization deviation sequence is extracted to represent the slight fluctuation of the message interval. The statistical distribution method is used for dynamic maintenance to obtain the range of variation of the deviation value, and it is applied to the quantitative evaluation of the reliability of the digital output of the instrument transformer in the smart substation.

[0030] Step S2 is divided into the following sub-steps: S2.1: For the time-series data of the current transformer digital message sequence and the reference pulse sequence that have already been fused using a unified time frame, the loss of the correspondence between their respective time-series data due to separate processing can be avoided. The timestamps of the current transformer digital message sequence and the reference pulse sequence are extracted from the unified time frame. The time corresponding to each frame of the current transformer digital message sequence in the first dimension of the unified time frame is traversed to obtain the timestamp of each frame and form a current transformer timestamp sequence; the time corresponding to each pulse of the reference pulse sequence in the second dimension of the unified time frame is traversed to obtain the timestamp of each pulse and form a reference timestamp sequence. During extraction, it is ensured that the position information of the current transformer timestamp sequence and the reference timestamp sequence corresponding to the same moment on the unified time frame remains in a one-to-one correspondence, thereby preserving the original time-series relationship and accurately reflecting the time offset of the current transformer digital message sequence relative to the reference pulse sequence, facilitating subsequent deviation calculation operations.

[0031] S2.2: By separating the current transformer timestamp sequence and the reference timestamp sequence from the unified time frame to form a one-to-one mapping, the difference between the timestamp sequences does not need to be quantized to account for the impact of average delay. Then, a synchronization deviation sequence is calculated based on the current transformer timestamp sequence and the reference timestamp sequence. Under the unified time frame, the difference between the elements in the current transformer timestamp sequence and the corresponding elements in the reference timestamp sequence is used as the element value of the synchronization deviation sequence. Each element in the synchronization deviation sequence represents the timing offset of the current transformer digital message sequence compared to the reference pulse sequence at that moment. The specific calculation steps of the synchronization deviation sequence are as follows: first, calculate the first term of the current transformer timestamp sequence and subtract the first term of the reference timestamp sequence to obtain the first difference of the synchronization deviation sequence; then, subtract the second term of the current transformer timestamp sequence and subtract the second term of the reference timestamp sequence to obtain the second difference of the synchronization deviation sequence, and so on, until the difference at all times has been calculated. Using the direct difference method for instantaneous offset quantization avoids the impact of accumulated errors and can capture the irregular discreteness of message intervals, directly reflecting the characteristics of jitter.

[0032] S2.3: A synchronization deviation sequence has been generated. Continuous recording after time-series offset is beneficial for dynamic tracking rather than static observation, adapting to the needs of long-term drift accumulation. The synchronization deviation sequence distribution is maintained in real time. An initial distribution container, such as an empirical distribution function, is used to accumulate and count the entire synchronization deviation sequence. After acquiring new unified time frame group data, the frequencies corresponding to each deviation interval in the distribution container are updated. During maintenance, only synchronization deviation sequence elements within a certain fixed window are retained, which directly reflect the current jitter situation. Furthermore, the statistical distribution is defined as a function mapping deviation values ​​to frequencies, facilitating subsequent step S3 to call this distribution to evaluate the phase walk trajectory, thereby achieving continuous quantification of deviation evolution and suppressing the adverse effects of hidden losses on the protection device.

[0033] In step S2, the synchronization deviation sequence and statistical distribution have been obtained periodically according to a unified time window and maintained in real time to measure the dynamic situation of the timing offset of the digital message sequence of the current transformer, providing data for subsequent drift trade-offs. There is no problem of missing interval jitter in the traditional method. Therefore, it can be continuously monitored in the smart substation, realizing continuous monitoring of synchronization reliability.

[0034] Based on the synchronization deviation sequence and the statistical distribution of the deviation sequence obtained in step S2, step S3 maps the synchronization deviation sequence to a phase walk trajectory and establishes the phase shift stacking quantity; at the same time, it processes the message interval sequence to obtain the pulse scattering rate, and then uses Gaussian process regression to fuse it into a drift tradeoff coefficient, thus analyzing the long-term drift and instantaneous jitter. The reason for this scheme is that the quantized synchronization deviation sequence can be mapped to a phase trajectory to obtain the accumulated offset, Mandel segmentation can reflect the fractal properties of the message interval sequence, and Gaussian process regression can fuse heterogeneous features into a single coefficient, so it can be applied to the multi-scale feature quantization of the drift assessment of the digital output of the instrument transformer in smart substations.

[0035] Step S3 is divided into the following sub-steps: S3.1: The synchronization deviation sequence represents a continuous time offset record, and the cyclical pattern of long-period drift cannot be accurately represented by a linear model. The synchronization deviation sequence is mapped to a phase wandering trajectory. A portion of the transformer timestamp sequences are selected from the synchronization deviation sequence. The statistical distribution deviation values ​​of each sequence are used as weights, and corresponding weights are assigned to the deviation values ​​of each sequence at the same time. The deviation values ​​are sequentially converted into phase values, and these are accumulated and moduloed to obtain the phase values ​​at each time point of the phase wandering trajectory. The phase value is obtained by taking the cumulative deviation value modulo 2π. The specific calculation method is as follows: starting from the first deviation value in the synchronization deviation sequence, the accumulated value is added to the cumulative value, and modulo 2π is taken to obtain the first phase value on the phase wandering trajectory. This is repeated for each subsequent deviation value until all values ​​in all synchronization deviation sequences are accumulated once, and then modulo 2π is taken to obtain the corresponding phase value, forming a complete phase wandering trajectory sequence. By accumulating and calculating the modulo, the deviation is mapped to a periodic phase space, thus avoiding the overflow of linear phase modulo accumulation. This more accurately reflects the long-term drift cycle characteristics of the deviation and provides a clearer description of the implicit cumulative effect of the transformer's digital output.

[0036] S3.2: The phase walk trajectory is a periodic sequence, which makes it easy to analyze and stack the offset using local windows without having to consider the problem of it being canceled out by transient effects under global average statistics. The phase shift stacking amount is obtained by calculating the cumulative offset within each sliding window: by using a fixed-length sliding window on the phase travel trajectory, the absolute value of the difference between two adjacent phase values ​​within the sliding window is accumulated to obtain the phase shift stacking amount sequence; taking a sliding window width of ten phase values ​​as an example, the accumulated absolute difference between the phase values ​​from the first to the tenth phase value of the phase travel trajectory is taken as the first phase shift stacking amount value; then the sliding window is shifted forward by one position (from the second to the eleventh phase value), and the accumulated absolute difference contained in the window is calculated to obtain the next phase shift stacking amount value; ... this shifting cycle continues until the sliding window superimposes the accumulated absolute differences between all adjacent phase values ​​contained in the entire phase travel trajectory to form the phase shift stacking amount sequence; by quantifying the local offset stacking through the accumulated absolute difference, the amplification effect of instantaneous jitter is reflected, the local dynamics of message interval jitter are reflected, and the sensitivity to short-term fluctuations in the digital output of the current transformer is increased.

[0037] S3.3: The message interval sequence is the adjacent difference between the timestamp sequences of the current transformers. Random delay information is mixed into the messages, facilitating the use of fractal analysis to expose irregularly discrete self-similar structures rather than simple linear measurements. The fractal dimension is used instead of a uniform metric, and the pulse scattering rate is obtained by using the Mandelbrot partitioning method to calculate the message intervals after calculating the dispersion. The time difference between adjacent timestamps obtained from the current transformer timestamp sequence is the message interval sequence. Then, the Mandelbrot partitioning method is used to iteratively divide the message interval sequence into multiple groups of self-similar segments. It is determined whether the segments satisfy the Mandelbrot set boundary condition (the boundary of a point set on the complex plane where the iterative function will not diverge). This boundary condition is applied to judge the fractal self-similarity of the segments. The dispersion is then calculated based on the local Hearst exponent of each segment, and the pulse scattering rate is obtained by taking the geometric mean of the Hearst exponents of each segment. For the message interval sequence, it is sequentially divided into two equal parts to test the self-similarity. If the message interval sequence does not meet the self-similarity condition, it is then sequentially divided into two equal parts. The sequence is divided into four parts, and the self-similarity of each part is checked until each self-similar sub-segment satisfies the Mandelbrot set boundary condition. Then, the Hearst exponent of each sub-segment is calculated by the rescaling range analysis method. The Hearst exponent is estimated by the slope of the fluctuation range and time span under the logarithmic scale. The Hearst exponent of the first sub-segment is multiplied by the Hearst exponent of the second sub-segment, and so on until the Hearst exponent of the last sub-segment is obtained. The product value is then taken as the base of the number of sub-segments to obtain the pulse scattering rate. The geometric mean is used to integrate the fractal dispersion to measure the random scattering of the interval sequence. The quantifiable interval sequence can better capture the scattering characteristics of instantaneous jitter and has higher estimation accuracy for the random delay of the digital output of the transformer.

[0038] S3.4: The phase shift stacking sequence and pulse scattering rate have been quantified as dual-scale features, which can more effectively fuse heterogeneous data under nonlinear relationships rather than simply ignoring complex dependencies through linear combinations. To obtain the drift tradeoff coefficients, we use Gaussian process regression. Specifically, the phase shift stacking sequence and pulse scattering rate are used as input features. After preprocessing and normalization, these input features are directly input into the Gaussian process regression model. A radial basis function kernel is used to fit the nonlinear mapping from features to output, thus obtaining the drift tradeoff coefficients. The calculation involves inverting the covariance matrix and predicting the mean, aiming to obtain the joint drift weight under linear combination of features. Specifically, each value in the phase shift stacking sequence is subtracted from its minimum value and divided by the difference between its maximum and minimum values ​​to obtain a value normalized to the 0-1 range. The pulse scattering rate is then normalized in the same way. The process involves obtaining normalized values ​​in the 0-1 interval to form the input feature vector. Gaussian process regression is then used, employing a radial basis function kernel to calculate the covariance matrix between input features. The radial basis function, i.e., the Gaussian kernel, is used to measure the similarity between feature points. The posterior distribution is obtained by mixing the inverse of the covariance matrix with the observation noise, and a unique single drift tradeoff coefficient is obtained through the posterior mean. Finally, a nonlinear mapping is used to fuse the features, ensuring that the drift tradeoff coefficient can fully capture the joint effect. This allows for better integration of information on long-term drift and instantaneous jitter, further enhancing the prediction effect on the overall synchronization reliability of the digital output of the transformer.

[0039] In step S3, a method is adopted to map the synchronization deviation sequence to a phase walk trajectory and establish the corresponding phase shift stacking amount. Then, based on the message interval sequence, multifractal analysis is performed to obtain the pulse scattering rate, and the drift tradeoff coefficient is obtained through Gaussian process regression fusion. This method realizes multi-scale quantitative analysis of the synchronization detection of the digital output of the instrument transformer. At the same time, by utilizing the periodic mapping of the phase walk trajectory and considering the long-term cumulative drift cycle mode of the phase-locked loop, the masking effect of linear deviation on implicit periodic fluctuations is offset, and the clock offset evolution of the electronic instrument transformer output message can be accurately distinguished in the smart substation environment. The sliding window accumulation method of the phase shift stacking amount can quantify the local amplification effect of instantaneous jitter, ensuring that the detection process is most sensitive to the uncertain random delay caused by sampling buffer queuing. This ensures the time stability index evaluation of the fault judgment sequence by the protection device. Whether the protection device can operate correctly and reliably is directly related to the delay of the message under test. This technology reduces the drift factor, thereby reducing the occurrence of false alarms or missed alarms caused by ultra-long messages.

[0040] In message interval processing, the message interval sequence is obtained using Mandelbrot segmentation (an iterative bisection method based on fractal self-similarity), and the Hearst exponent is calculated as a dispersion index to obtain the pulse scatter rate. This method can accurately reflect the fractal structure and long-range correlation of the interval sequence. Compared with traditional statistical variance analysis, this method can better distinguish between persistent trends caused by temperature drift or aging and irregular random jitter. It can be used for the detection of irregular discreteness under independent clocks of current transformers, thereby suppressing erroneous tripping caused by the implicit offset of the direction-finding threshold due to jitter in the measurement message interval caused by this factor.

[0041] Gaussian process regression, as an integrated module, maps the different qualitative characteristics of phase shift stacking and pulse scattering rate to the same drift tradeoff coefficient. Through nonlinear kernel function mapping, it completes the mapping calculation between these two parameter values. The resulting drift tradeoff coefficient can be used for low-sample drift trend prediction through uncertainty modeling within a Bayesian framework, without requiring a specific parameter distribution. This is particularly beneficial for the specific optimization of the digital output of the instrument transformer, enabling real-time aggregation and fusion of its long-term drift and instantaneous jitter information, real-time correction of the synchronization stability threshold, and forming a closed-loop nested operation of the inspection process (collection, analysis, and intervention). This improves inspection efficiency while eliminating the impact of network congestion or external disturbances on the digital output of the instrument transformer. In summary, the entire process can utilize two-scale feature extraction and predictive fusion to more robustly detect the synchronization characteristics of the optical fiber transmission messages of the instrument transformer. To a certain extent, it avoids the bias problem of detection conclusions caused by single-frame data anomalies in closed-loop testing, enhancing the advantage of the continuity between frames of protection information received by the protection device from the instrument transformer.

[0042] The drift tradeoff coefficient is injected into the synchronization stability domain, and a deviation weight vector is constructed through the drift tradeoff coefficient. The threshold is adjusted according to this vector to achieve nonlinear regulation of the synchronization deviation. The reason for using this method is that the drift tradeoff coefficient has integrated dual-scale characteristics, and a large-scale weight vector that can correspond to any complex distribution can be generated through non-additive injection. In addition, the dynamic threshold adjustment is also based on the generalized entropy calculation on the correction boundary. Therefore, it does not consider the case where non-Gaussian drift is easily ignored by linear methods. It can optimize jitter in real time during actual operation of the substation and ensure the accuracy of the timing of the protection device.

[0043] The drift tradeoff coefficients are obtained by using the Gaussian process regression output obtained in step S3 and injected into the synchronization stability domain. Then, the threshold is dynamically adjusted according to the weight vector to obtain a nonlinear adaptive intelligent substation instrument transformer digital output synchronization intervention method.

[0044] Step S4 is divided into the following sub-steps: S4.1: The drift tradeoff coefficient quantifies the heterogeneous jitter characteristics, enabling the nonlinear injection vector space to extract scale invariance, while nonlinear methods ignore complex correlations; the drift tradeoff coefficient is injected into the synchronous stable domain to construct the deviation weight vector.

[0045] The synchronization stability region is the number of dimensions of the vector space corresponding to the statistical distribution interval of a synchronization deviation sequence. The bias weight vector is obtained by using the drift tradeoff coefficient as the exponent of the initial values ​​of each dimension. Each element of the bias weight vector is the power-law drift tradeoff coefficient corresponding to the initial value of the dimension of the synchronization stability region. This process involves using the first initial value of the synchronization stability region as the base and the drift tradeoff coefficient as the exponent to perform a power operation to obtain the first element of the bias weight vector; using the second initial value of the synchronization stability region as the base and the drift tradeoff coefficient as the exponent to perform a power operation to obtain the second element of the bias weight vector; and so on, processing all initial values ​​of dimensions to obtain all bias weight vectors. Then, power-law injection is used to achieve nonlinear amplification, which better captures the scale invariance of the drift and thus better captures the non-Gaussian distribution characteristics of the digital output of the transformer, obtaining a more accurate weight representation of the synchronization deviation. For example, the calculation formula can be: [Formula omitted for brevity]. ,in The first element of the bias weight vector represents the first element. One element, The first stable region represents the synchronous stable region. Initial values ​​for each dimension (preset as interval normalized values) This represents the drift tradeoff coefficient.

[0046] S4.2: The deviation weight vector has formed a nonlinear structure and is suitable for generalized entropy transformation to measure dispersion, in order to adapt to the distribution emphasis at the tail, while non-standard entropy cannot take into account non-additive effects; dynamic threshold adjustment is performed based on the deviation weight vector. The specific approach is as follows: The deviation weight vector is processed using the Tsallis entropy form, i.e., the generalized entropy form, to describe the partial information uncertainty of the non-additive system. The transformation after raising the normalized vector elements to the power of the preset non-additive parameter is used as the new threshold. The actual operation steps are as follows: First, the sum of all elements of the deviation weight vector is calculated as the denominator for normalization. Then, each element value in the deviation weight vector is divided by its sum to obtain its normalized value, and the power of each element is raised to the power of the preset non-additive parameter. Finally, all powers are summed to obtain an intermediate sum. This intermediate sum is then subtracted from one, and the result is divided by one and the preset non-additive parameter is subtracted to obtain the new threshold. Simultaneously, the generalized entropy transformation quantifies the non-additive dispersion to better handle the dynamic boundary changes caused by complex drifts, improving the sensitivity to the jitter changes in the digital output of the transformer and enhancing the robustness of the intervention mechanism. For example, the Tsallis entropy expression can be used to transform each element of the deviation weight vector, and the power of each normalized component can be raised to the power of the sum to obtain the new threshold. For example, the calculation formula could be: applying the Tsallis entropy form to the bias weight vector to calculate the normalized vector elements. The result of summing the powers of a factor and then transforming it serves as the new threshold; the calculation formula is as follows: ,in This indicates the adjusted threshold. The first element of the bias weight vector represents the first element. One element, This indicates a preset non-additive parameter (with a value greater than 1 to emphasize the tail distribution). This represents the total number of vector dimensions. This represents the sum of all elements in the bias weight vector.

[0047] In step S4, the deviation weight vector is generated as a drift tradeoff coefficient. Based on this, a corresponding threshold is generated to nonlinearly adjust the jitter of the digital output of the current transformer. On this basis, the optimized boundary setting for triggering intervention is realized, thereby alleviating the problem that linear methods cannot be applied to smart substations.

[0048] When the deviation weight vector triggers the phase-locked loop adjustment command, it returns to S1 to execute the closed-loop nested process of data acquisition verification and threshold iteration. The reason for choosing this scheme is that the deviation weight vector represents the current time drift state. When its value changes to the threshold, the minimum adjustment amount will be directly sent to the phase-locked loop. At the same time, new data points will be collected to update the threshold. This avoids the problem of not being able to respond in time due to timeout when using a static threshold. It achieves the purpose of correcting clock drift in real time during substation operation, reduces the inspection frequency, and reduces the losses caused by hidden losses.

[0049] Based on the deviation weight vector formed by S4 and the dynamic adjustment threshold, it is determined whether the limit has been exceeded, and an adjustment command is issued. The process then loops back to S1 to continue data acquisition, verification, and threshold iteration, providing a closed-loop intervention method for maintaining the synchronization of the digital output of the instrument transformer in the smart substation.

[0050] Step S5 is divided into the following sub-steps: S5.1: Its deviation weight vector has a non-linear weighting term, which highlights the extreme deviation through Holder norm quantification. Therefore, it is more sensitive to the average deviation than the Euclidean norm. It is sufficient to track and monitor whether the deviation weight vector reaches the threshold.

[0051] The deviation weight vector norm and the dynamically adjusted threshold are compared in real time. If the norm is greater than the threshold, it is determined that a threshold has been reached. The process of calculating the norm is to take the first element of the deviation weight vector and raise it to the pre-hold exponent, take the second element of the deviation weight vector and raise it to the pre-hold exponent, and so on until the last element of the deviation weight vector is raised to the pre-hold exponent. The exponents are added together and the sum is taken as the root of the pre-hold exponent to obtain the deviation weight vector norm. The pre-hold exponent is set to be greater than 1, so that the magnitude of the endpoint value is given more importance. The entire vector amplitude is quantified by the pre-hold norm. The non-Euclidean distance is used to find the position where the numerical fluctuation reaches the warning value, so as to accurately locate the starting position of the current transformer digital output jitter threshold exceeding the threshold and ensure accurate trigger response.

[0052] S5.2: Based on the touch judgment, it is beneficial to obtain the minimum increment based on the maximum element, so as to achieve fine adjustment instead of average adjustment and ignore local peaks; if the trigger level is too low at this time, the phase-locked loop adjustment command will be triggered.

[0053] The minimum adjustment amount is used as the adjustment command to enter the phase-locked loop inside the transformer. The maximum element of the deviation weight vector is used as the basis for calculating the minimum adjustment amount. The calculation method is as follows: find the maximum value in the deviation weight vector, then multiply the maximum value by a pre-set scaling factor, with a value between (0,1). This scaling factor ensures that the adjustment amplitude is as small as possible. The final adjustment is completed by scaling by the maximum value. To quantify the time offset, this ensures that the adjustment amount and the offset of the time base jitter of the measured signal have an order-of-magnitude correspondence, effectively suppressing the instantaneous amplification of the digital output jitter of the transformer, thereby making the timing judgment of the protection device more accurate and stable.

[0054] S5.3: The adjustment command is sent to the substation monitoring system, and the iterative threshold is gradually modified according to the resulting symbol to adapt to the changing situation rather than ignoring the directionality of the deviation. After the adjustment command is issued, the system returns to step S1 to cyclically verify the newly collected data and continue iterating the threshold.

[0055] Step S1 is executed repeatedly to obtain a new unified time frame. When verifying the adjustment effect of the new synchronization deviation sequence using the aforementioned method, if the deviation decreases, the iteration threshold is adjusted upwards; otherwise, it is adjusted downwards. The specific calculation formula is as follows: First, calculate the average value of the new synchronization deviation sequence and subtract the average value of the old synchronization deviation sequence to obtain the average difference. Then, calculate the value of the sign function. The sign function is calculated as follows: if the average difference is greater than 0, output +1; if the average difference is less than 0, output -1; if the average difference is equal to 0, output 0. Multiply the sign function by the preset iteration step size as an intermediate factor. The iteration step size is a small positive value given in advance. The intermediate factor value is accumulated to obtain the intermediate factor. Finally, the current dynamic adjustment threshold is multiplied by the intermediate factor to obtain the threshold after iteration. The sign function is used to guide the adaptive iterative optimization of the boundary value. Finally, the synchronization calibration of the digital output of the transformer is achieved, reducing false tripping.

[0056] After the above operations, when the deviation weight vector trigger level reaches the threshold, the phase-locked loop is activated for adjustment, and sampling verification and threshold iteration are performed to correct the jitter of the digital output of the current transformer in a closed loop, so as to maintain continuous synchronization in the smart substation and prevent false tripping caused by drift amplification.

[0057] Example 2: Figure 2 The present invention provides a digital output synchronization detection system for a current transformer, comprising: a data acquisition module, a deviation maintenance module, a drift fusion module, a threshold dynamic module, and an adjustment cycle module; The data capture module synchronously captures the digital messages and reference pulses of the current transformer under independent external time base constraints and records them to a unified time frame; The deviation maintenance module uses a unified time frame to obtain a synchronization deviation sequence; The drift fusion module maps the synchronization deviation sequence to a phase walk trajectory, and constructs the phase shift stacking amount by accumulating the offset through a sliding window. It combines the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generates the drift tradeoff coefficient through Gaussian process regression. The threshold dynamic module injects the drift tradeoff coefficient into the synchronization stability region to form a deviation weight vector, and dynamically adjusts the threshold accordingly. The adjustment loop module is used to trigger the phase-locked loop adjustment command when the deviation weight vector reaches the threshold, and then return to the data interception module to complete the acquisition verification and threshold iteration.

[0058] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0059] It should be noted that the system of the present invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting a variety of hardware environments and usage requirements.

[0060] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

[0061] It should be noted that, in this document, the use of relational terms such as "first" and "second" is merely to distinguish one entity or operation from another, and does not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0062] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for detecting the synchronization of digital output of a current transformer, characterized in that, Including the following steps: S1: Synchronously intercept the digital messages and reference pulses of the current transformer under independent external time base constraints and record them to a unified time frame; S2: Obtain the synchronization deviation sequence using a unified time frame; S3: Map the synchronization deviation sequence to a phase walk trajectory, and construct the phase shift stacking amount by accumulating the offset through a sliding window. Combine this with the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generate the drift tradeoff coefficient through Gaussian process regression. S4: Inject the drift tradeoff coefficient into the synchronization stability region to form a deviation weight vector, and dynamically adjust the threshold accordingly; S5: When the deviation weight vector reaches the threshold, the phase-locked loop adjustment command is triggered, and the process returns to step S1 to complete the acquisition verification and threshold iteration.

2. The method for detecting the synchronization of digital output of a current transformer according to claim 1, characterized in that: Step S1 includes deploying an independent external time base as a data acquisition constraint source, connecting the acquisition device between the current transformer output port and the reference pulse source, generating a trigger signal from the independent external time base to initiate the current transformer digital message sequence and reference pulse sequence interception operation, buffering the current frame and timestamp and the current pulse and timestamp at the beginning of each cycle to ensure that the interception window covers the complete message interval cycle, recording the sequence to a unified time frame, and using the independent external time base cycle as the unit of a two-dimensional structure, the first dimension stores the current transformer digital message sequence frame content and timestamp pair, and the second dimension stores the reference pulse sequence pulse content and timestamp pair, merging the two sequences by aligning the timestamps.

3. The method for detecting the synchronization of digital output of a current transformer according to claim 2, characterized in that: Step S2 includes extracting the transformer timestamp sequence and the reference timestamp sequence from the unified time frame, calculating the difference between the elements of the transformer timestamp sequence and the elements of the reference timestamp sequence to form a synchronization deviation sequence, initializing the distribution container to accumulate and count the synchronization deviation sequence, updating the frequency of each deviation interval in the distribution container when a new set of unified time frame data is received, using a rolling update mechanism to retain the synchronization deviation sequence elements within the most recent fixed window, and defining the statistical distribution using the deviation value to frequency mapping function.

4. The method for detecting the synchronization of digital output of a current transformer according to claim 3, characterized in that: Step S3 includes projecting the phase walk trajectory from the synchronization deviation sequence, converting the elements of the synchronization deviation sequence into phase values ​​using statistical distribution weights, forming a phase walk trajectory sequence by taking the cumulative deviation modulo 2π, applying a sliding window to calculate the absolute difference between adjacent elements on the phase walk trajectory to construct a phase shift stacking sequence, calculating the adjacent differences from the transformer timestamp sequence to form a message interval sequence, performing Mandel Segmentation on the message interval sequence to divide it into self-similar segments, calculating the Hearst exponent of each segment and taking the geometric mean as the pulse scattering rate.

5. The method for detecting the synchronization of digital output of a current transformer according to claim 4, characterized in that: The phase shift stacking sequence and pulse scattering rate are normalized and then input into the Gaussian process regression model. Drift tradeoff coefficients are generated by fitting the radial basis function kernel.

6. The method for detecting the synchronization of digital output of a current transformer according to claim 5, characterized in that: Step S4 includes defining the synchronization stability region as a vector space of the number of statistical distribution intervals corresponding to the synchronization deviation sequence, and applying the drift tradeoff coefficient as a power exponent to the initial value of each dimension to form a deviation weight vector.

7. The method for detecting the synchronization of digital output of a current transformer according to claim 6, characterized in that: The normalized elements of the bias weight vector are calculated by applying the Tsallis entropy form, and the summation of the preset non-additive parameters is used as the adjustment threshold.

8. The method for detecting the synchronization of digital output of a current transformer according to claim 7, characterized in that: Step S5 includes real-time comparison of the deviation weight vector norm and the dynamic adjustment threshold. The magnitude of the deviation weight vector is quantified by the Holder norm to determine if it is triggered. The minimum adjustment amount based on the maximum element of the deviation weight vector is generated as a phase-locked loop adjustment command and sent to the internal phase-locked loop of the transformer. Step S1 is executed repeatedly to obtain a new unified time frame. Steps S2 to S4 are applied to the new synchronization deviation sequence to verify the adjustment effect. The dynamic adjustment threshold is guided iteratively by the sign function of the difference between the average value of the new synchronization deviation sequence and the average value of the old synchronization deviation sequence.

9. A current transformer digital output synchronization detection system, used to implement the current transformer digital output synchronization detection method according to any one of claims 1-8, characterized in that, It includes a data capture module, a deviation maintenance module, a drift fusion module, a threshold dynamic module, and an adjustment loop module; among which, The data capture module synchronously captures the digital messages and reference pulses of the current transformer under independent external time base constraints and records them to a unified time frame; The deviation maintenance module uses a unified time frame to obtain a synchronization deviation sequence; The drift fusion module maps the synchronization deviation sequence to a phase walk trajectory, and constructs the phase shift stacking amount by accumulating the offset through a sliding window. It combines the pulse scattering rate obtained by performing Mandel segmentation on the message interval and calculating the discreteness, and generates the drift tradeoff coefficient through Gaussian process regression. The threshold dynamic module injects the drift tradeoff coefficient into the synchronization stability region to form a deviation weight vector, and dynamically adjusts the threshold accordingly. The adjustment loop module is used to trigger the phase-locked loop adjustment command when the deviation weight vector reaches the threshold, and then return to the data interception module to complete the acquisition verification and threshold iteration.

Images