Method for taming a rubidium atomic clock based on a sliding window dynamic compensation and related product
By using a sliding window dynamic compensation method, the discipline adjustment cycle of the rubidium atomic clock is shortened. Incremental control is performed using the change in frequency deviation, which solves the problem of long adjustment cycle in long-wavelength discipline methods and improves the discipline accuracy and stability of the rubidium atomic clock.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NAT TIME SERVICE CENT CHINESE ACAD OF SCI
- Filing Date
- 2026-03-31
- Publication Date
- 2026-06-05
AI Technical Summary
Existing long-wavelength training methods suffer from insufficient training accuracy and stability because the adjustment period is too long and they cannot respond promptly to short-term frequency drifts of rubidium atomic clocks.
A sliding window dynamic compensation method is adopted, which divides the sampling period into a first time period and a second time period. The change in frequency deviation is used to compensate for the short time window data, and the adjustment amount is calculated by the least squares method to shorten the discipline adjustment period. A rubidium atomic clock time difference model is constructed for dynamic compensation.
It achieves the shortening of the discipline adjustment cycle, improves the discipline accuracy and stability of rubidium atomic clocks, timely tracks and compensates for short-term frequency drift, and avoids over-adjustment and controls oscillations while maintaining the anti-interference capability of long-wave signals.
Smart Images

Figure CN122151468A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of time and frequency synchronization technology, specifically to a method and related products for taming a rubidium atomic clock based on dynamic compensation using a sliding window. Background Technology
[0002] Rubidium atomic clocks have been widely used in many fields such as communications, navigation, and power synchronization due to their good short-term stability and low cost. However, rubidium atomic clocks have an inherent frequency drift (i.e., aging) phenomenon. As the operating time increases, their frequency accuracy gradually decreases, requiring periodic calibration using an external, highly stable reference signal, i.e., discipline. Currently, a common discipline method is to use the pulse-per-second (1PPS) signal output by satellite navigation systems (such as BeiDou and GPS) as a reference. This method can effectively correct the frequency deviation of the rubidium clock. However, satellite signals are susceptible to factors such as obstruction, interference, multipath effects, and satellite clock bias, limiting their reliability in complex environments and making it difficult to meet the requirements of high-availability applications.
[0003] Long-wave timing signals (such as the Loland-C system) have advantages such as long propagation distance, strong anti-interference capability, and wide coverage, making them an important backup or supplementary means for satellite timing. However, long-wave signals are susceptible to ionospheric changes, superposition of ground waves and sky waves, multipath effects, and propagation path delay fluctuations during propagation. The periodic interference in their timing data, especially the celestial periodic component, results in poor stability of the directly measured time difference data, making it difficult to directly use for the discipline control of high-precision atomic clocks. Existing long-wave discipline methods typically use a 24-hour data window for parameter estimation and adjustment to suppress the influence of celestial periodic interference. However, this type of method has a long adjustment cycle (usually 1 day), causing the rubidium atomic clock to be in a free-running state for a long time between adjacent adjustment intervals. This makes it impossible to respond promptly to the short-term frequency drift of the rubidium clock itself, limiting further improvements in discipline accuracy and stability.
[0004] Therefore, how to shorten the training and adjustment cycle and improve the stability of rubidium atomic clocks has become a technical problem that urgently needs to be solved in this field. Summary of the Invention
[0005] The purpose of this invention is to provide a method and related products for taming a rubidium atomic clock based on dynamic compensation using a sliding window, so as to overcome the problem that the traditional long-wavelength taming method in the prior art has insufficient taming accuracy and stability due to the excessively long taming adjustment cycle.
[0006] The present invention solves the above-mentioned technical problems through the following technical solution: This invention provides a method for dynamically compensating and taming a rubidium atomic clock based on a sliding window, comprising the following steps: Step 1: Set the number of iterations q=1. Divide the sampling period into a first duration and a second duration according to a preset ratio. Sequentially acquire the longwave-rubidium clock time difference data D between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. Step 2: Based on long-wavelength-rubidium clock time difference data D q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. q Determine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; Step 3: Retain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; based on the adjusted rubidium atomic clock, obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration. ; Step 4: Based on frequency deviation B q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to step two.
[0007] A further improvement of this invention lies in the fact that the rubidium atomic clock time difference model is specifically as follows:
[0008] in, This represents the quadratic trend term, used to characterize the frequency drift of a rubidium clock. This represents the phase difference between the rubidium atomic clock and the long-wave timing signal at the initial moment; This indicates the deviation between the initial frequency of the rubidium atomic clock and the standard frequency. Indicates the frequency drift coefficient; This represents a multi-periodic term, used to characterize the impact of external periodic interference on the time difference signal of a long-wavelength signal; where M represents the number of periodic components in the multi-periodic term; and i represents the index variable for summing the multi-periodic term. This represents the coefficient of the cosine component in the i-th periodic term; This represents the coefficient of the sinusoidal component in the i-th periodic term; This represents the frequency of the i-th periodic term; The noise term is represented, specifically including measurement noise and electromagnetic interference; t is the time vector.
[0009] A further improvement of the present invention is that the frequency deviation B q Change Specifically:
[0010] Where A is the frequency deviation of the qth iteration, and B is the frequency deviation of the qth iteration. q F represents the frequency deviation B of the (q-1)th iteration. q-1 .
[0011] A further improvement of this invention is that when q=1, F=0.
[0012] A further improvement of this invention lies in that, based on frequency deviation B... q Changes in longwave-rubidium clock time difference data Compensation will be provided, specifically as follows: in, This is the compensated long-wavelength-rubidium clock time difference data; The i-th long-wavelength-rubidium clock time difference data before compensation; The time for the i-th data point; This is a compensation item.
[0013] A further improvement of the present invention is that the sampling period is 24 hours and the preset ratio is 1:23.
[0014] The present invention also provides a system for taming a rubidium atomic clock based on dynamic compensation using a sliding window, comprising: The first module is used to set the number of cycles q=1, divide the sampling period into a first duration and a second duration according to a preset ratio, and sequentially acquire the long-wave-rubidium clock time difference data D between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. The second module is used for D based on long-wavelength-rubidium clock time difference data.q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. q Determine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; The third module is used to store the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; and to obtain the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration based on the adjusted rubidium atomic clock. ; The fourth module is used for frequency deviation B. q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to the second module to continue the loop.
[0015] The present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method for taming a rubidium atomic clock based on sliding window dynamic compensation as described above.
[0016] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method described above for dynamically compensating and disciplining a rubidium atomic clock based on a sliding window.
[0017] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described above for dynamically compensating and disciplining a rubidium atomic clock based on a sliding window.
[0018] Compared with the prior art, the positive and progressive effects of the present invention are as follows: The present invention provides a method for taming a rubidium atomic clock based on dynamic compensation using a sliding window. By dividing the sampling period into a first duration and a second duration, and introducing a sliding window and a dynamic compensation mechanism, the method compensates for the short-term window data by utilizing the change in frequency deviation and then splices it with the long-term window data. This achieves the shortening of the taming adjustment period to the first duration without weakening the anti-interference capability of long-wavelength signals (maintaining the 24-hour data window to suppress daily periodic interference). It can track and compensate for the short-term frequency drift of the rubidium atomic clock in a timely manner, thereby improving the taming accuracy and stability of the rubidium atomic clock.
[0019] Furthermore, by using the change in frequency deviation as an adjustment amount, incremental control of the rubidium atomic clock is achieved, avoiding over-adjustment and control oscillations caused by single measurement errors or model fitting deviations. This makes the taming process smoother and more stable, which is beneficial to improving the long-term stability of the rubidium atomic clock's output frequency. Attached Figure Description
[0020] The accompanying drawings are provided to further understand the invention and constitute a part of this invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.
[0021] Figure 1 Schematic diagram of the improved algorithm for taming rubidium atomic clocks for long waves; Figure 2 The improved algorithm for taming rubidium atomic clocks for long-wavelength taming yields the clock difference curves between the rubidium clock and UTC (NTSC). Figure 3 Improved algorithm for taming rubidium atomic clocks for long waves - stability comparison chart of tamed rubidium atomic clocks. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.
[0023] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.
[0024] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.
[0025] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.
[0026] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."
[0027] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. This is an explanation of the present invention and not a limitation thereof.
[0028] To address the technical problem that existing long-wavelength training methods have long adjustment cycles and cannot respond promptly to short-term frequency drift of rubidium atomic clocks, thus limiting training accuracy and stability, this invention provides a method for training rubidium atomic clocks based on sliding window dynamic compensation, comprising the following steps: Step 1: Set the number of iterations q=1. Divide the sampling period into a first duration and a second duration according to a preset ratio. Sequentially acquire the longwave-rubidium clock time difference data D between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. Step 2: Based on long-wavelength-rubidium clock time difference data D q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. q Determine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; Step 3: Retain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; based on the adjusted rubidium atomic clock, obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration. ; Step 4: Based on frequency deviation B q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to step two.
[0029] In step one, the sampling period is the total time window used to collect data, for example, it can be set to 24 hours. Based on the characteristics of periodic interference in long-wave signals (such as the sky periodic component generated by the superposition of sky waves and ground waves), this sampling period is divided into two parts: a first duration and a second duration. The first duration is a shorter time window used for rapid adjustment, for example, 1 hour; the second duration is the remaining time window used to obtain stable time difference data unaffected by short-term adjustments, for example, 23 hours. Using a long-wave timing receiver and a rubidium atomic clock, 1PPS signals output from both are collected respectively. The time difference between these signals is measured using a time interval counter, sequentially obtaining the time difference data corresponding to the first and second durations. These data together constitute a complete time difference data sequence, denoted as Dq. Simultaneously, to accurately describe the behavior of the rubidium atomic clock under long-wave signal interference, a mathematical model needs to be established based on the periodic fluctuation characteristics of the long-wave signal, namely the rubidium atomic clock time difference model, for subsequent parameter estimation.
[0030] In step two, the time difference data Dq obtained in step one is substituted into the constructed rubidium atomic clock time difference model. Since the model contains multiple undetermined parameters, the least squares method is used to optimally estimate these parameters, thereby calculating the phase difference and frequency deviation Bq of the current rubidium atomic clock relative to the long-wavelength reference signal. Then, the adjustment amount is determined based on the number of iterations q. If q = 1, i.e., the first iteration, there is no historical adjustment information, and the calculated phase difference and frequency deviation Bq are directly used as the adjustment amount, simultaneously used to correct the phase and frequency of the rubidium atomic clock. If q is not 1, it indicates that at least one adjustment has already been performed. In this case, only the frequency deviation Bq calculated this time needs to be compared with the frequency deviation Bq-1 of the previous iteration to calculate the change in frequency deviation, and this change is used as the adjustment amount for this iteration. Based on the determined adjustment amount, the rubidium atomic clock is corrected, resulting in an adjusted rubidium atomic clock, which then proceeds to step three for a new round of compensation and adjustment.
[0031] In step three, the portion of longwave-rubidium clock time difference data collected in the current cycle corresponding to the second duration (i.e., a longer time window, such as 23 hours) is retained. This data is considered relatively stable and less affected by the adjustment process, representing historical data. Then, using the adjusted rubidium atomic clock as a reference, the next sampling cycle begins. Within the first duration of the new sampling cycle (i.e., a shorter time window, such as 1 hour), the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by this adjusted rubidium atomic clock is acquired again. .
[0032] In step four, to utilize a shorter time window for parameter estimation in the next loop, it's necessary to address the data discontinuity issue caused by the adjustment operation. This step uses the frequency deviation change calculated in step two to adjust the new time difference data obtained in step three, based on the adjusted rubidium atomic clock. Compensation is performed. The purpose of compensation is to restore this new data to the reference baseline before adjustment, ensuring that it is physically consistent with the previously retained second-duration historical data. The data obtained after compensation is denoted as... ; Use the new data after this compensation Replace the old data portion corresponding to the first duration in the previous round of time difference data Dq. Then, use the replaced data... The time difference data from the second time interval retained in step three is concatenated in chronological order to form a complete set of time difference data for the next round of parameter estimation. Finally, the concatenated new data is used as the input Dq for the next round, and the process returns to step two to begin a new round of adjustment loop.
[0033] The present invention provides a method for taming a rubidium atomic clock based on dynamic compensation using a sliding window. By introducing a sliding window mechanism, the traditional long-period (e.g., 24-hour) adjustment window is decomposed into a short window (first duration) and a long window (second duration). After each adjustment, the newly acquired short window data is used, and compensation is applied based on the current frequency deviation change. This short window data is then combined with the previous long window data to form a new sliding window with the same data length but containing the latest adjustment information. In this way, the latest short-term data is utilized in each cycle, while long-term historical data ensures the stability of parameter estimation. Through this dynamic compensation and sliding window approach, long-term interference such as the day cycle of long-wave signals is effectively suppressed, while the adjustment period is shortened from the original 24 hours to the first duration (e.g., 1 hour), thereby enabling timely tracking and compensation for short-term frequency drift of the rubidium atomic clock. By implementing the above steps, this method can shorten the discipline adjustment cycle of the rubidium atomic clock while maintaining the advantage of strong anti-interference capability of long-wave timing signals, thereby improving the short-term stability and tracking accuracy of the rubidium atomic clock and solving the problem in the prior art that the long adjustment cycle makes it impossible to respond to the drift of the rubidium clock itself in a timely manner.
[0034] Specifically, the rubidium atomic clock time difference model is as follows:
[0035] in, This represents the quadratic trend term, used to characterize the frequency drift of a rubidium clock. This represents the phase difference between the rubidium atomic clock and the long-wave timing signal at the initial moment; This indicates the deviation between the initial frequency of the rubidium atomic clock and the standard frequency. Indicates the frequency drift coefficient; This represents a multi-periodic term, used to characterize the impact of external periodic interference on the time difference signal of a long-wavelength signal; where M represents the number of periodic components in the multi-periodic term; and i represents the index variable for summing the multi-periodic term. This represents the coefficient of the cosine component in the i-th periodic term; This represents the coefficient of the sinusoidal component in the i-th periodic term; This represents the frequency of the i-th periodic term; The noise term is represented, specifically including measurement noise and electromagnetic interference; t is the time vector.
[0036] The phase, initial frequency deviation, and frequency drift (aging) characteristics of the rubidium atomic clock are described by a quadratic trend term, which forms the basis of discipline control. Simultaneously, by introducing a multi-period term, inherent and regular interferences in long-wavelength signals (especially the daily periodic component) are separated from the time difference data, preventing these interferences from contaminating the frequency deviation estimation. A noise term is used to handle the remaining random errors, making the model more closely resemble the actual physical process. Using this model for least-squares fitting enables more accurate and robust extraction of the true state parameters (phase difference and frequency deviation) of the rubidium atomic clock from the disturbed time difference data, providing a reliable basis for subsequent high-precision adjustments, thereby improving the overall accuracy and anti-interference capability of the discipline system.
[0037] Specifically, frequency deviation B q Change Specifically:
[0038] Where A is the frequency deviation of the qth iteration, and B is the frequency deviation of the qth iteration. q F represents the frequency deviation B of the (q-1)th iteration. q-1 .
[0039] Specifically, when q=1, F=0.
[0040] Using the change in frequency deviation, rather than the absolute deviation, as the adjustment amount essentially achieves incremental control. Its physical meaning is that each adjustment only corrects the frequency drift that has newly occurred since the last adjustment. This aligns with the characteristic that the frequency drift of a rubidium atomic clock is a continuous accumulation process. Incremental adjustment avoids over-adjustment and oscillations caused by measurement noise or model errors, making the discipline process smoother and more stable, thereby improving the long-term stability of the rubidium atomic clock's output frequency.
[0041] Specifically, based on frequency deviation B q Changes in longwave-rubidium clock time difference data Compensation will be provided, specifically as follows: in, This is the compensated long-wavelength-rubidium clock time difference data; The i-th long-wavelength-rubidium clock time difference data before compensation; The time for the i-th data point; This is a compensation item.
[0042] Adjusting a rubidium atomic clock is achieved by changing its frequency. This frequency change leads to a linearly cumulative deviation in subsequent time difference measurements. The compensation formula utilizes this frequency deviation change to construct an inverse linear term to counteract the deviation introduced by the adjustment. This allows the new and old data segments to be stitched together under the same reference standard. Through precise linear compensation, the consistency of data within the sliding window in terms of physical meaning and time reference is ensured. This ensures that the next parameter estimation (least squares method) based on this data accurately reflects the true state of the rubidium atomic clock, which is a key technical guarantee for achieving window sliding and shortening the adjustment cycle.
[0043] Specifically, the sampling period is 24 hours, and the preset ratio is 1:23.
[0044] By employing a 24-hour sampling period and a 1:23 ratio, the interference from the day cycle of long-wave signals was effectively suppressed, and the discipline adjustment period was successfully shortened to 1 hour, thereby improving the discipline accuracy and short-term stability of the rubidium atomic clock. Specifically, the total sampling period was set to 24 hours, which matches the period of the most significant periodic interference (day cycle) in the long-wave signal, effectively suppressing its impact on parameter estimation. Simultaneously, the first duration was set to 1 hour, and the second duration to 23 hours. This means that in each cycle, there is a 1-hour "new" data window used to capture the latest frequency changes of the rubidium atomic clock, while the 23-hour window is the "old" data window used to ensure the stability of parameter estimation and the ability to suppress long-cycle interference. For example, the first cycle collects time difference data from 0-1 hour and 1-24 hours; the second cycle collects new data from 25-26 hours, compensates for it, and then combines it with the previously retained 1-24 hour data to form a new 2-26 hour data window, and so on. This setup reduces the adjustment cycle from 24 hours to 1 hour while ensuring that the data window length for parameter estimation remains at 24 hours, thus balancing rapid response and anti-interference capabilities.
[0045] The present invention also provides a system for taming a rubidium atomic clock based on dynamic compensation using a sliding window, comprising: The first module is used to set the number of cycles q=1, divide the sampling period into a first duration and a second duration according to a preset ratio, and sequentially acquire the long-wave-rubidium clock time difference data D between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. The second module is used for D based on long-wavelength-rubidium clock time difference data. q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. qDetermine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; The third module is used to store the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; and to obtain the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration based on the adjusted rubidium atomic clock. ; The fourth module is used for frequency deviation B. q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to the second module to continue the loop.
[0046] The system for taming a rubidium atomic clock based on sliding window dynamic compensation provided by this invention uses a first module for data acquisition, a second module for core control and parameter estimation, and a third and fourth module to collaboratively update and compensate the data in the sliding window. The four modules work in a loop to form a closed-loop control, thereby automatically and continuously taming the rubidium atomic clock. It can automatically and efficiently execute the sliding window dynamic compensation taming process, effectively shorten the taming adjustment cycle, and improve the taming accuracy and stability of the rubidium atomic clock.
[0047] In a specific embodiment of the present invention, see Figure 1 A method for taming a rubidium atomic clock based on sliding window dynamic compensation includes the following steps: System construction and data acquisition: The time difference between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock is measured using a multi-channel time interval counter to obtain the raw long-wave-rubidium clock time difference data; An improved rubidium clock time difference model is established: Considering the periodic fluctuation characteristics in long-wave signals, a periodic term is added to the traditional quadratic polynomial rubidium clock time difference model to construct an improved time difference model:
[0048] In the formula, This represents the quadratic trend term, used to characterize the frequency drift of a rubidium clock; where, This represents the phase difference between the rubidium atomic clock and the long-wave timing signal at the initial moment; This indicates the deviation between the initial frequency of the rubidium atomic clock and the standard frequency. Indicates the frequency drift coefficient; This represents a multi-periodic term, used to characterize the impact of external periodic interference on the time difference signal of a long-wavelength signal; where M represents the number of periodic components in the multi-periodic term; and i represents the index variable for summing the multi-periodic term. This represents the coefficient of the cosine component in the i-th periodic term; This represents the coefficient of the sinusoidal component in the i-th periodic term; This represents the frequency of the i-th periodic term; The noise term is represented, specifically including measurement noise and electromagnetic interference; t is the time vector.
[0049] Initial Adjustment and Data Retention: Due to the large initial phase and frequency deviations of the rubidium clock, a large-value adjustment is required initially. Afterward, the phase and frequency deviations of the rubidium clock can be controlled simply by adjusting the frequency (the phase of the rubidium clock should not be adjusted frequently, as this can easily cause sawtooth fluctuations in the 1PPS output signal of the rubidium clock).
[0050] Initial adjustment: Based on the first 24 hours of long-wave-rubidium clock time difference data (denoted as...). The initial phase difference of the rubidium clock was estimated using the least squares method. and frequency offset (Frequency deviation B1) Perform the first coarse adjustment of the rubidium clock phase and frequency. After the first adjustment, the change in rubidium clock frequency is: ( = ); Data retention: Retain the last 23 hours of long-wave-rubidium clock time difference data after the initial adjustment (denoted as...). A total of 82,800 data points were collected and used as the historical baseline for subsequent sliding windows. Retaining 23 hours of historical data is to allow for the rapid formation of a new 24-hour window by supplementing it with new 1-hour data points, avoiding the problem of excessively long adjustment cycles caused by repeatedly collecting 24-hour data.
[0051] (1) The first sliding adjustment process is as follows: it includes new data acquisition and frequency offset compensation.
[0052] New data acquisition: After the initial adjustment, continuously acquire new long-wave-rubidium clock time difference data (denoted as ) according to the set sliding period (e.g., 1 hour). (A total of 3600 data points).
[0053] Frequency offset compensation: Since the initial adjustment has changed the rubidium clock frequency offset Newly collected data Compared with historical data There is a systematic offset in the cumulative time difference caused by frequency deviation. To ensure data continuity, it is necessary to... Compensation will be provided. in, This is the i-th long-wavelength-rubidium clock time difference data after the first iteration compensation; This refers to the i-th long-wavelength-rubidium clock time difference data before the first iteration compensation; The time for the i-th data point (in seconds); This is a compensation term for the first iteration, used to offset the impact of the initial frequency offset adjustment on the new data, ensuring that the new data is consistent with the time base of historical data.
[0054] Sliding window data reorganization: This involves recompacting the new data. Compared with the 23-hour historical data retained By combining the data, a new 24-hour data window can be created: This window contains both 23 hours of historical data (to maintain the capture of statistical characteristics of long-wave periodic interference) and 1 hour of new data (to reflect the latest drift status of the rubidium clock), achieving the dual characteristics of full-cycle coverage and real-time updates.
[0055] Iterative estimation of adjustment amount based on sliding window: New frequency offset estimation: based on the new 24-hour time difference data after recombination. The current frequency offset of the rubidium clock (denoted as ) is then estimated using the least squares method. ); New adjustment calculation: The new frequency offset minus the previous frequency offset is used as the frequency offset adjustment for this time and applied to the rubidium clock. That is, the latest adjustment is: ; Historical data update: Update the retained historical data as follows (That is, discard the earliest 1 hour of data and keep the latest 23 hours of data) to prepare for the next sliding window.
[0056] (2) The second sliding adjustment process is as follows: it includes new data acquisition and frequency offset compensation.
[0057] New data acquisition: After the initial sliding adjustment, based on the adjusted rubidium atomic clock, new long-wavelength-rubidium clock time difference data (denoted as ) is continuously acquired according to a set sliding period (e.g., 1 hour). (A total of 3600 data points).
[0058] Frequency offset compensation: Based on the previous sliding adjustment that changed the rubidium clock frequency offset. Newly collected data Compared with historical data There is a systematic offset in the cumulative time difference caused by frequency deviation. To ensure data continuity, it is necessary to... Compensation will be provided. in, This is the i-th long-wavelength-rubidium clock time difference data after the second iteration compensation; This refers to the i-th long-wavelength-rubidium clock time difference data before the second iteration compensation; The time for the i-th data point (in seconds); This is a compensation term for the second iteration, used to offset the impact of the previous frequency offset adjustment on the new data, ensuring that the new data is consistent with the time base of historical data.
[0059] Sliding window data reorganization: This involves recompacting the new data. With retained historical data By combining the data, a new 24-hour data window can be created: This window contains both 23 hours of historical data (maintaining the capture of statistical characteristics of long-wave periodic interference) and 1 hour of new data (reflecting the latest drift status of the rubidium clock), continuing to maintain the dual characteristics of full-cycle coverage and real-time updates.
[0060] Iterative estimation of adjustment amount based on sliding window: New frequency offset estimation: based on the new 24-hour time difference data after recombination. The current frequency offset of the rubidium clock (denoted as ) is then estimated using the least squares method. ); New adjustment calculation: The new frequency offset minus the previous frequency offset is used as the frequency offset adjustment for this time and applied to the rubidium clock. That is, the latest adjustment is: ; Historical data update: Update the retained historical data as follows (That is, discard the earliest 1 hour of data and keep the latest 23 hours of data) to prepare for the next sliding window.
[0061] Cyclic execution: Subsequently, new data is continuously collected, compensated, window reorganized, estimated and adjusted according to the set sliding step size of 1 hour to achieve high-frequency closed-loop control of the rubidium clock.
[0062] Through the above strategy, the rubidium clock adjustment cycle is shortened from 1 day to 1 hour (this embodiment uses a 1-hour step size, but other sliding step sizes of several hours can also be set according to the actual project technical requirements). This maintains the ability of 24-hour data to suppress long-wave periodic interference, and can also respond to the frequency drift characteristics of the rubidium clock in real time through high-frequency adjustment, breaking through the limitation of the original scheme of free operation on a single day, and improving the discipline accuracy and stability.
[0063] The method of this invention, while retaining the ability to suppress interference such as the 24-hour long-wave data cycle, significantly shortens the rubidium clock adjustment cycle from the traditional 24 hours to several hours (1 hour in this embodiment) or even shorter, thereby improving the system's response speed to short-term frequency drift of the rubidium clock and avoiding instability caused by the free operation of the rubidium clock on a single day. By using a sliding window combined with a frequency offset compensation strategy, the continuity of the data window in time is ensured, overcoming the problem of time base jumps that may be caused by segmented data processing.
[0064] The time difference data between the UTC (NTSC) (National Time Service Center of Chinese Academy of Sciences) reference 1PPS signal and the rubidium clock 1PPS was collected as a performance evaluation benchmark.
[0065] See Figure 2 This involves using MATLAB to statistically analyze the time difference waveforms between a rubidium atomic clock and UTC (NTSC); see [link / reference]. Figure 3 This refers to the waveform of the Allan variance (used to measure the stability of the rubidium atomic clock) calculated using MATLAB for the time difference data between the rubidium atomic clock and UTC (NTSC). Figure 3 The statistical results of the stability of the rubidium atomic clock are shown in Table 1. The experimental results show that after adopting this method, the standard deviation of the clock difference between the tamed rubidium clock and UTC (NTSC) is 15.61 ns, and the long-term stability (100,000 seconds) of the tamed rubidium clock is improved from 1.86e-12 to 5.85e-13. The system control accuracy and stability are significantly improved. Table 1. Comparison of stability between free and docile states of a rubidium clock.
[0066] Based on the same inventive concept, this application provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of a method for dynamically compensating and disciplining a rubidium atomic clock based on a sliding window. The memory may include main memory, such as high-speed random access memory, or it may also include non-volatile memory, such as at least one disk storage device. The processor, network interface, and memory are interconnected via an internal bus, which may be an industry-standard architecture bus, a peripheral component interconnection standard bus, an extended industry-standard architecture bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory stores the program; specifically, the program may include program code, which includes computer operation instructions. The memory may include main memory and non-volatile memory, and provides instructions and data to the processor.
[0067] Based on the same inventive concept, embodiments of this application provide a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the steps of the method for dynamically compensating and disciplining a rubidium atomic clock based on a sliding window. Specifically, the computer-readable storage medium includes, but is not limited to, volatile memory and / or non-volatile memory. The volatile memory may include RAM (Random Access Memory) and / or cache memory, etc. The non-volatile memory may include ROM (Read-Only Memory), hard disk, flash memory, optical disk, magnetic disk, etc.
[0068] Based on the same inventive concept, this application provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions that, when executed by a computer device, cause the computer device to perform the steps of the above-described method for dynamically compensating and disciplining a rubidium atomic clock based on a sliding window.
[0069] Those skilled in the art will understand that embodiments of the present invention can be provided as methods or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM (Compact Disc Read-Only Memory), optical storage, etc.) containing computer-usable program code.
[0070] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer apparatus or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0071] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer device or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0072] These computer program instructions may also be loaded onto a computer device or other programmable data processing equipment to cause a series of operational steps to be performed on the computer device or other programmable equipment to produce a process implemented by the computer device, thereby providing instructions that execute on the computer device or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0073] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0074] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for taming a rubidium atomic clock based on dynamic compensation using a sliding window, characterized in that, Includes the following steps: Step 1: Set the number of iterations q=1. Divide the sampling period into a first duration and a second duration according to a preset ratio. Sequentially acquire the longwave-rubidium clock time difference data D between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. Step 2: Based on long-wavelength-rubidium clock time difference data D q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. q Determine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; Step 3: Retain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; based on the adjusted rubidium atomic clock, obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration. ; Step 4: Based on frequency deviation B q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to step two.
2. The method for taming a rubidium atomic clock based on dynamic compensation using a sliding window according to claim 1, characterized in that, The specific time difference model of the rubidium atomic clock is as follows: in, This represents the quadratic trend term, used to characterize the frequency drift of a rubidium clock. This represents the phase difference between the rubidium atomic clock and the long-wave timing signal at the initial moment; This indicates the deviation between the initial frequency of the rubidium atomic clock and the standard frequency. Indicates the frequency drift coefficient; This represents a multi-periodic term, used to characterize the impact of external periodic interference on the time difference signal of a long-wavelength signal; where M represents the number of periodic components in the multi-periodic term; and i represents the index variable for summing the multi-periodic term. This represents the coefficient of the cosine component in the i-th periodic term; This represents the coefficient of the sinusoidal component in the i-th periodic term; This represents the frequency of the i-th periodic term; The noise term is represented, specifically including measurement noise and electromagnetic interference; t is the time vector.
3. The method for taming a rubidium atomic clock based on dynamic compensation using a sliding window according to claim 1, characterized in that, Frequency deviation B q Change Specifically: Where A is the frequency deviation of the qth iteration, and B is the frequency deviation of the qth iteration. q F represents the frequency deviation B of the (q-1)th iteration. q-1 .
4. The method for taming a rubidium atomic clock based on dynamic compensation using a sliding window according to claim 1, characterized in that, When q=1, F=0.
5. The method for taming a rubidium atomic clock based on dynamic compensation using a sliding window according to claim 1, characterized in that, Based on frequency deviation B q Changes in longwave-rubidium clock time difference data Compensation will be provided, specifically as follows: in, This is the compensated long-wavelength-rubidium clock time difference data; The i-th long-wavelength-rubidium clock time difference data before compensation; The time for the i-th data point; This is a compensation item.
6. The method for taming a rubidium atomic clock based on dynamic compensation using a sliding window according to claim 1, characterized in that, The sampling period is 24 hours, and the preset ratio is 1:
23.
7. A system for taming a rubidium atomic clock based on sliding window dynamic compensation, characterized in that, include: The first module is used to set the number of cycles q=1, divide the sampling period into a first duration and a second duration according to a preset ratio, and sequentially acquire the long-wave-rubidium clock time difference data D between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first and second durations. q Based on the periodic fluctuation characteristics of long-wavelength signals, a time difference model for rubidium atomic clocks is constructed. The second module is used for D based on long-wavelength-rubidium clock time difference data. q Using a rubidium atomic clock time difference model, the phase difference and frequency deviation B of the rubidium atomic clock are calculated using the least squares method. q Determine if the current loop count is 1. If the result is yes, then set the phase difference and frequency deviation B. q As an adjustment, if the judgment result is negative, then the frequency deviation B will be adjusted. q Change As an adjustment amount; The rubidium atomic clock is adjusted based on the adjustment amount to obtain the adjusted rubidium atomic clock; The third module is used to store the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration; and to obtain the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the first duration based on the adjusted rubidium atomic clock. ; The fourth module is used for frequency deviation B. q Long-wave-rubidium clock time difference data Compensation is performed to obtain the longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first compensation duration. The longwave-rubidium clock time difference data between the 1PPS signal output by the longwave timing receiver and the 1PPS signal output by the rubidium atomic clock at the first time duration after sequential splicing and compensation. Given the long-wave-rubidium clock time difference data between the 1PPS signal output by the long-wave timing receiver and the 1PPS signal output by the rubidium atomic clock under the current second duration, let q = q + 1 to obtain the long-wave-rubidium clock time difference data D. q Then return to the second module to continue the loop.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method for taming a rubidium atomic clock based on dynamic compensation using a sliding window as described in any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for taming a rubidium atomic clock based on dynamic compensation using a sliding window as described in any one of claims 1 to 6.
10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the steps of the method for taming a rubidium atomic clock based on dynamic compensation using a sliding window, as described in any one of claims 1 to 6.