A high-precision calibration method and system for substation clocks

Abstract

This invention belongs to the field of data processing technology, specifically relating to a high-precision calibration method and system for substation clocks. The method includes: real-time analysis of the timing signal; calculation of an instantaneous channel disturbance index reflecting interference intensity and an optimized instantaneous channel disturbance index distinguishing interference types; and dynamically generating an applicable pruning threshold that adaptively matches the channel conditions based on these indices; applying this dynamic threshold to the Viterbi decoding process. This invention can intelligently relax pruning conditions to protect the correct path when strong interference occurs, and tighten conditions to improve efficiency when the channel is good, thereby enhancing the clock calibration's anti-interference capability and decoding accuracy in harsh electromagnetic environments such as substations.

Description

Technical Field

[0001] This invention relates to the field of data processing technology. More specifically, this invention relates to a high-precision calibration method and system for substation clocks. Background Technology

[0002] In modern intelligent substations, a unified high-precision time reference for the entire station is fundamental for advanced applications such as equipment status monitoring and relay protection. The synchronization accuracy of the time synchronization system directly affects the safe and stable operation of the power grid. Currently, substations mainly use wired hard time synchronization based on IRIG-B codes or wireless time synchronization based on GPS or BeiDou. However, substations are complex environments with strong electromagnetic interference. High-voltage equipment generates instantaneous, high-intensity electromagnetic pulses during operation. These interferences are superimposed on the transmission link of the time synchronization signal, causing signal distortion.

[0003] To accurately recover time information from noisy signals, existing technologies generally employ Viterbi decoding algorithms based on the maximum likelihood criterion. This algorithm approximates the original information sequence by searching and retaining paths with optimal path metrics in the decoding grid graph, and controls computational complexity by discarding less likely paths through a pruning strategy.

[0004] However, the error correction capability of the Viterbi algorithm is closely related to the constraint length of the convolutional code. The constraint length is usually a fixed parameter selected during the design phase after weighing the error correction capability and computational resources. A smaller constraint length means a smaller state space for the decoding algorithm, faster computation speed, and lower resource consumption, but limited error correction capability, making it unable to effectively cope with strong pulse interference generated by events such as switching operations in substations. Using a larger constraint length can improve anti-interference performance, but the decoding complexity is proportional to the number of states and grows exponentially. The computational overhead may cause the decoding delay to exceed the real-time requirements of microsecond-level synchronization of the timing system. Therefore, a fixed constraint length prevents the timing system from dynamically adjusting its error correction capability and computational overhead according to real-time changes in the channel environment. Ultimately, this may lead to calibration failure due to insufficient performance in harsh environments, or unnecessary resource waste due to over-design in favorable environments. Summary of the Invention

[0005] To address the aforementioned technical problem of poor clock calibration performance in substations, this invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a high-precision calibration method for a substation clock, comprising:

[0007] A timing signal containing interference pulses is acquired and digitized to obtain a received symbol sequence. A window is established along the received symbol sequence, and the instantaneous channel disturbance index at each moment is obtained. The instantaneous channel disturbance index is positively correlated with the sum of squares of the deviations of symbol amplitudes exceeding the reference amplitude within the window, and negatively correlated with the window size. An optimized instantaneous channel disturbance index at each moment is obtained. The optimized instantaneous channel disturbance index is positively correlated with the square of the largest single amplitude deviation within the window, and negatively correlated with the product of the instantaneous channel disturbance index and the window size. The optimized instantaneous channel disturbance index is multiplied by a basic pruning threshold to obtain an applicable pruning threshold. The applicable pruning threshold is used to perform Viterbi decoding on the received symbol sequence and to calibrate the substation clock.

[0008] This invention dynamically generates an applicable pruning threshold that matches the current channel conditions by calculating the instantaneous channel disturbance index, which reflects the deviation of signal amplitude, and the optimized instantaneous channel disturbance index, which characterizes the concentration of interference energy, in real time. When strong interference occurs, the pruning conditions can be relaxed to protect the correct decoding path from being incorrectly eliminated, while the conditions are tightened to improve decoding efficiency when the channel is good. This significantly enhances the anti-interference capability and decoding accuracy of the clock calibration system in strong electromagnetic interference environments such as substations.

[0009] Preferably, the creation of the window includes:

[0010] Using any given time as a reference, select W-1 times immediately preceding the given time, and construct a window of the given time with the given time.

[0011] Preferably, the instantaneous channel disturbance index satisfies the expression:

[0012] ;

[0013] In the formula, Represents the instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This indicates the preset average received signal reference amplitude; This represents the unit step function.

[0014] This invention introduces a unit step function, which allows the accumulation of received symbols whose amplitudes exceed the normal range. This enables the calculation to focus on high-energy pulse interference caused by disconnector operation, while ignoring amplitude reduction caused by normal signal fading or low-level noise. This allows for more accurate identification of interference events that pose a substantial threat to decoding, thus improving the perturbation index's relevance and effectiveness.

[0015] Preferably, the establishment of the unit step function includes:

[0016] The unit step function is derived from express;

[0017] when hour, ;when hour, .

[0018] Preferably, the optimized instantaneous channel disturbance index satisfies the expression:

[0019] ;

[0020] In the formula, Let represent the optimized instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This indicates the preset average received signal reference amplitude; It is a unit step function; Represents the instantaneous channel disturbance index at time t; Represents the normalization function; This is a local minimum value, used to avoid a denominator of 0.

[0021] This invention constructs an optimized index that can effectively measure energy concentration by calculating the ratio of the maximum single amplitude deviation energy to the total deviation energy within a window. It can identify pulse interference spikes with highly concentrated energy, avoid misjudging normal signal symbol switching as interference, thereby improving the accuracy of channel state assessment and providing a more reliable basis for subsequent dynamic pruning.

[0022] Preferably, the Viterbi decoding of the received symbol sequence using an applicable pruning threshold includes: when updating the surviving paths of the state at time t of the decoding trellis, retaining the best path with the optimal path metric; for other candidate paths, if the difference between their path metric and the metric of the best path is less than or equal to the applicable pruning threshold at time t, then the candidate path is also retained as a surviving path.

[0023] The dynamic retention strategy of this invention achieves an intelligent balance between fault tolerance at critical moments and high efficiency at normal moments, thereby improving the robustness of decoding.

[0024] Preferably, the timing signal is a BPSK modulated BPC signal.

[0025] Preferably, the step of acquiring and digitizing the timing signal includes: acquiring an analog timing signal using an analog front-end (AFE); and sampling the analog timing signal using an analog-to-digital converter (ADC) at a preset sampling frequency to obtain a received symbol sequence.

[0026] This invention uses an analog front-end (AFE) and an analog-to-digital converter (ADC) for high-frequency sampling, ensuring high-fidelity capture of the timing signal waveform and the characteristics of the narrow pulse interference superimposed on it, providing a high-quality data foundation for subsequent accurate calculation of amplitude deviation and energy concentration.

[0027] Preferably, the sampling frequency is 2.5MHz.

[0028] Secondly, the present invention provides a high-precision calibration system for a substation clock, comprising a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned high-precision calibration method for a substation clock is implemented.

[0029] By adopting the above technical solution, a computer program for a high-precision calibration method for substation clocks is generated and stored in a memory for loading and execution by a processor. Terminal devices are then manufactured based on the memory and processor for convenient use.

[0030] The beneficial effects of this invention are as follows: By adaptively calculating the pruning threshold for each decoding window, it can effectively cope with the signal decoding requirements under different electromagnetic interference environments. In traditional methods, while the fixed design of the constraint length can improve anti-interference capability, it also leads to a significant increase in computational complexity and resource consumption. Especially in high-precision time synchronization systems, this design can cause decoding delays to exceed the requirements of real-time synchronization. However, by dynamically adjusting the pruning threshold, the system can flexibly control the complexity of the decoding process based on the real-time assessment of channel interference characteristics, thereby improving decoding accuracy in strong interference environments and saving computational resources in good environments, avoiding the waste caused by over-design. This adaptive adjustment strategy not only improves the overall performance of the system but also ensures that the time synchronization system remains efficient and reliable in various environments, guaranteeing high-precision calibration of substation clocks and effectively balancing the contradiction between decoding performance and resource consumption. Attached Figure Description

[0031] Figure 1 This is a flowchart illustrating a high-precision calibration method for a substation clock according to the present invention;

[0032] Figure 2 This is an illustrative comparison of CPU usage before and after dynamically adjusting the pruning threshold;

[0033] Figure 3This is a schematic diagram showing the comparison of execution efficiency before and after dynamically adjusting the pruning threshold. Detailed Implementation

[0034] This invention discloses a high-precision calibration method for substation clocks, referring to... Figure 1 This includes steps S1-S3:

[0035] S1: Acquire and preprocess the timing signal to obtain a digitized sequence of received symbols.

[0036] It should be noted that the substation environment is filled with high-intensity, pulsed electromagnetic interference generated by disconnecting switches, circuit breaker operations, etc., which poses a serious threat to time synchronization signals transmitted wirelessly or via power line carrier. Standard time synchronization signals, such as BPSK-modulated BPC signals, are highly susceptible to decoding errors in this environment, leading to decreased clock accuracy or even loss of synchronization. Therefore, this invention first captures the original time synchronization signal through a high-fidelity front-end circuit and digitizes it, providing reliable data input for subsequent anti-interference decoding algorithms.

[0037] Specifically, the timing signal is acquired and preprocessed, including: using an analog front-end (AFE) module to acquire the timing signal, and then using an analog-to-digital converter (ADC) device to convert the analog timing signal at a preset sampling frequency to obtain a digitized received symbol sequence for analysis in subsequent steps.

[0038] The sampling frequency needs to be much higher than the symbol rate of the timing signal to accurately capture the signal waveform and details of any interference pulses that may be superimposed on it. For example, the sampling frequency can be set to 2.5MHz. The received symbol sequence is a complex sequence containing amplitude and phase information for each sampling moment.

[0039] S2: Based on the amplitude characteristics of the received symbols, obtain the instantaneous channel disturbance index at each moment; based on the amplitude change characteristics of the received symbols, and combined with the instantaneous channel disturbance index at each moment, obtain the optimized instantaneous channel disturbance index at each moment; combined with the optimized instantaneous channel disturbance index at each moment and the basic pruning threshold, obtain the applicable pruning threshold at each moment.

[0040] It should be noted that while the traditional Viterbi decoding algorithm performs well in handling Gaussian white noise channels, it has certain shortcomings when dealing with the strong pulse-type interference unique to substations. A brief, strong interference pulse can occur at any moment in the decoding lattice diagram, causing the metric of a path containing correct symbols to deteriorate rapidly, even falling below that of other incorrect paths. This leads to the correct path being prematurely pruned and eliminated by the standard Viterbi algorithm, resulting in decoding failure. Therefore, this invention considers dynamically adjusting the Viterbi algorithm's pruning process, assessing the channel's interference intensity in real time. When strong pulse interference is detected, the pruning conditions are dynamically and temporarily relaxed, increasing the pruning threshold to allow multiple candidate correct paths with similar path metrics to coexist, thereby protecting the correct path from safely passing through the interference area. Conversely, when the channel is good, strict pruning conditions are maintained, and the pruning threshold is reduced to ensure decoding efficiency.

[0041] It should be noted that strong interference in substations mainly manifests as short-duration, high-amplitude pulses. Traditional indicators such as signal-to-noise ratio (SNR) are based on long-term statistical averages and cannot effectively capture such sudden interference events. Effective indicators should focus on identifying instantaneous and dramatic changes in signal amplitude. When a strong interference pulse is superimposed on the timing signal, it causes the amplitude of the received symbol to deviate instantly and significantly from its position on the ideal modulation constellation diagram. By statistically analyzing the severity of this deviation, the occurrence of strong interference can be accurately identified. Therefore, this invention constructs an instantaneous channel disturbance index that can characterize the intensity of pulse interference in real time by establishing a sliding window and statistically analyzing the severity of the received symbol amplitude exceeding the normal range within the sliding window. Specifically, the greater the deviation of the received symbol amplitude from the expected received symbol amplitude at each moment within the sliding window, the greater the probability of strong pulse interference in the channel at that moment, and the larger the instantaneous channel disturbance index at that moment.

[0042] Specifically, based on the amplitude characteristics of the received symbols, the instantaneous channel disturbance index at each moment is obtained, including:

[0043] Using any given time as a reference, select W-1 times immediately preceding that time, and construct a window of those times with the given time. For example, W=500.

[0044] The instantaneous channel disturbance index satisfies the following expression:

[0045] ;

[0046] In the formula, Represents the instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This represents the preset average received signal reference amplitude, for example, depending on the receiver hardware design. It can be set to 100; Represents the unit step function, when hour, =1, hour, The value is 0. The purpose of the unit step function is to make the calculation of the instantaneous channel disturbance index only focus on those samples whose amplitude exceeds the normal range, while ignoring the amplitude reduction caused by signal fading, so as to specifically calculate the impulse interference.

[0047] In the formula, This represents the difference between the amplitude of the i-th received symbol in the window at time t and the preset average received signal reference amplitude. This represents the square of the difference. The larger the value, the greater the deviation of the received symbol within the sliding window at time t from the expected received symbol amplitude, indicating a higher probability of strong pulse interference in the channel at time t, and a larger instantaneous channel disturbance index at time t; where The larger the value, the more samples exceeding the preset average received signal reference amplitude are within the sliding window at time t, indicating a greater likelihood of strong pulse interference in the channel at time t. The larger the instantaneous channel disturbance index at time t, the greater the confidence level.

[0048] For example, For time t, if within the sliding window of 500 samples, 498 samples have amplitudes around 1.0, for example, between 0.9 and 1.1, their contribution to the formula is 0. However, two of these samples are subject to strong interference, with amplitudes of... and Then the instantaneous channel disturbance index at that moment is: .

[0049] It should be noted that during normal symbol switching of the timing signal, such as transitioning from a low level representing "0" to a high level representing "1", amplitude fluctuations will occur due to the band-limiting effect of the filters in the communication system. The instantaneous amplitude of this normal signal transition fluctuation may exceed the reference amplitude. This can lead to misidentification as interference, resulting in a misjudgment of channel conditions. Considering that the core difference between real interference fluctuations and normal transition fluctuations lies in the concentration of energy, real interference pulses have their energy highly concentrated within a very short time, manifesting as extreme spikes in amplitude at one or two sampling points within the sliding window. In contrast, the energy of normal transition fluctuations is relatively dispersed throughout the entire symbol transition time, exhibiting a common, gradual change in amplitude at multiple consecutive sampling points. Therefore, if the sum of amplitude deviations of received symbols at all times within the sliding window corresponds to a given moment, the greater the contribution of the maximum amplitude deviation of received symbols at all times within the sliding window, the greater the likelihood of a real interference spike at that moment, and consequently, the larger the optimized instantaneous channel disturbance index at that moment. Therefore, this invention distinguishes between these two types of fluctuations by calculating the ratio of the maximum single-point energy contribution to the total energy within the window, constructing an optimized index that accurately reflects the level of real interference.

[0050] Preferably, based on the amplitude variation characteristics of the received symbols and combined with the instantaneous channel disturbance index at each time moment, the optimized instantaneous channel disturbance index at each time moment is obtained, including:

[0051] The optimized instantaneous channel disturbance index satisfies the following expression:

[0052] ;

[0053] In the formula, Let represent the optimized instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This indicates the preset average received signal reference amplitude; It is a unit step function; Represents the instantaneous channel disturbance index at time t; Represents the normalization function; This is a minimum value used to avoid a denominator of 0. For example, .

[0054] In the formula, This means that the unit step function value at time t is used as the weight to weight the square of the difference between the i-th received symbol in the window at time t and the preset average received signal reference amplitude; the numerator term The term represents the square of the largest single sample amplitude deviation within the window at time t, and represents the peak energy within the window; the denominator term... This represents the sum of the squares of all amplitude deviations within the window, signifying the total energy within the window. The larger the value, the greater the sum of the amplitude deviations of the received symbols at all times within the sliding window corresponding to time t. The greater the contribution of the maximum amplitude deviation of the received symbols at all times within the sliding window, the greater the possibility of a real interference spike at time t. Therefore, the larger the optimized instantaneous channel disturbance index at time t will be.

[0055] For example, if the total deviation energy within the window is 13, and the largest single deviation energy comes from the sample with an amplitude of 4.0, its energy is... Then the ratio before normalization is: Assuming that the maximum calculated ratio is 0.95 and the minimum is 0.05 across all times, the optimized instantaneous channel disturbance index at time t using the maximum-minimum normalization method is: The It is obtained by retaining three decimal places.

[0056] It should be noted that the computational complexity of Viterbi decoding is directly related to the number of surviving paths it retains. A fixed pruning strategy cannot account for channel variations; it may be too strict during interference, resulting in the loss of correct paths, or too lenient during normal operation, wasting computational resources. An ideal strategy should directly correlate the leniency of pruning with the actual interference level. A larger instantaneous channel disturbance index indicates a greater probability of a real interference spike at that moment, thus requiring a larger adaptive pruning threshold after the decoding window is established. This allows the algorithm to retain more candidate paths, effectively preventing correct paths from being incorrectly excluded due to temporary metric disadvantages. Therefore, this invention adaptively adjusts the basic pruning threshold based on the optimized instantaneous channel disturbance index.

[0057] Preferably, the applicable pruning threshold for each time moment is obtained by combining the optimized instantaneous channel disturbance index at each time moment with the basic pruning threshold, including:

[0058] The applicable pruning threshold at each time point satisfies the following expression:

[0059] ;

[0060] In the formula, This represents the applicable pruning threshold at time t; Let represent the optimized instantaneous channel disturbance index at time t; This represents the basic pruning threshold, for example. .

[0061] For example, .

[0062] In contrast, at a time when the channel is good, It may be close to 0, for example Then the pruning threshold at that moment is The pruning conditions are more stringent.

[0063] At this point, the applicable pruning thresholds for each time point have been obtained.

[0064] S3: Apply the appropriate pruning threshold to complete Viterbi decoding and output the calibrated time information.

[0065] It should be noted that applying the applicable pruning threshold to the Viterbi algorithm, through this improvement, allows the decoder to dynamically retain more candidate paths when strong interference is detected. This is equivalent to increasing the robustness and fault tolerance of decoding at critical moments, while returning to an efficient decoding mode during channel recovery. Ultimately, it can more reliably recover the correct time information from the interference and complete the high-precision calibration of the local clock. Figure 2 To show the comparison of CPU usage before and after dynamically adjusting the pruning threshold, such as... Figure 3 The comparison charts show the CPU utilization and execution efficiency before and after dynamically adjusting the pruning threshold, respectively.

[0066] Specifically, the application uses a pruning threshold to complete Viterbi decoding and implement clock calibration, including:

[0067] At time t when decoding the mesh graph, when updating the surviving path of a certain state, first, according to the standard Viterbi algorithm, compare all the preceding paths pointing to that state, find the path with the optimal metric, and record its metric value. For all other preceding paths pointing to this state, obtain their path metrics. If the path metric difference satisfies Then the candidate path It was also preserved as a survival path; if If the candidate path is not found, it is pruned and discarded. After the entire decoding process is completed, the path with the best global path metric is selected from all the paths that survive to the final moment as the final decoding result. High-reliability time information is extracted from the decoding result and used to calibrate the local clock of the substation terminal equipment.

[0068] For example, at time t, the decoder needs to update the state. Survival path: For moments of strong interference, as in the example above ; Measure of the optimal leader path There exists a candidate path, whose metric is... ; Measurement difference is ;because The candidate path is retained; the decoder will temporarily maintain two surviving paths in this state. For periods of good channel conditions, as in the example above... ; Measure of the optimal leader path There are identical candidate paths, and their metrics are... The measurement difference remains the same. ;because The candidate path is pruned; in this state, the decoder retains only the best surviving path.

[0069] This completes the high-precision calibration of the substation clock.

[0070] This invention also discloses a high-precision calibration system for substation clocks, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a high-precision calibration method for substation clocks according to the present invention.

[0071] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0072] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A high-precision calibration method for substation clocks, characterized in that, include: The timing signal containing the interference pulse is acquired and digitized to obtain the received symbol sequence; A window is established along the received symbol sequence to obtain the instantaneous channel disturbance index at each moment: ； In the formula, Represents the instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This indicates the preset average received signal reference amplitude; Represents the unit step function; The optimized instantaneous channel disturbance index is obtained at each time point. The optimized instantaneous channel disturbance index is positively correlated with the square of the largest single amplitude deviation within the window and negatively correlated with the product of the instantaneous channel disturbance index and the window size. The optimized instantaneous channel disturbance index is multiplied by the basic pruning threshold to obtain the applicable pruning threshold. Viterbi decoding of the received symbol sequence using the applicable pruning threshold includes: when updating the surviving path of the state at time t of decoding the mesh graph, retaining the best path with the optimal path metric. For other candidate paths, if the difference between their path metric and the metric of the best path is less than or equal to the applicable pruning threshold at time t, then the candidate path is also retained as a surviving path. Calibrate the substation clock.

2. The high-precision calibration method for a substation clock according to claim 1, characterized in that, The establishment window includes: Using any given time as a reference, select W-1 times immediately preceding the given time, and construct a window of the given time with the given time.

3. The high-precision calibration method for a substation clock according to claim 1, characterized in that, The establishment of the unit step function includes: The unit step function is derived from express; when hour, ;when hour, .

4. The high-precision calibration method for a substation clock according to claim 1, characterized in that, The optimized instantaneous channel disturbance index satisfies the expression: ； In the formula, Let represent the optimized instantaneous channel disturbance index at time t; This represents the window size at time t; This represents the i-th received symbol in the window at time t. express The amplitude; This indicates the preset average received signal reference amplitude; It is a unit step function; Represents the instantaneous channel disturbance index at time t; Represents the normalization function; This is a local minimum value, used to avoid a denominator of 0.

5. A high-precision calibration method for a substation clock according to claim 1, characterized in that, The timing signal is a BPSK modulated BPC signal.

6. A high-precision calibration method for a substation clock according to claim 1, characterized in that, The step of acquiring and digitizing the timing signal containing interference pulses includes: acquiring the analog timing signal using an analog front-end (AFE); and sampling the analog timing signal using an analog-to-digital converter (ADC) at a preset sampling frequency to obtain a received symbol sequence.

7. A high-precision calibration method for a substation clock according to claim 6, characterized in that, The sampling frequency is 2.5MHz.

8. A high-precision calibration system for substation clocks, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a high-precision calibration method for a substation clock according to any one of claims 1-7.

Images