Atomic clock group change-oriented elastic time scale generation method, system and equipment

Abstract

The invention relates to an atomic clock group change-oriented elastic time scale generation method, system and device. The method comprises the following steps: constructing an elastic factor graph model generated by time scales, generating a main time scale and a standby time scale which are consistent, carrying out time difference change accumulation of the main time scale and the standby time scale by utilizing atomic clock group change, compensating time difference and frequency deviation jump values of the time scales under the clock group change, generating an elastic time scale, and calculating the elastic time scale according to the elastic factor graph model. The continuity and stability of the time scale are ensured, and the monitoring performance of the atomic clock is improved. Meanwhile, by utilizing a plug-and-play architecture of the factor graph, factor nodes can be dynamically added or removed to update the model when the clock group changes, and the whole elastic factor graph model does not need to be reconstructed, so that the generation of the elastic time scale has the advantages of flexibility and expandability.

Description

Technical Field

[0001] The present application relates to the fields of time frequency and signal processing technology, and in particular to a method, system and device for generating an elastic time scale for changes in an atomic clock group. Background Art

[0002] Time-scaling algorithms play a vital role in timekeeping laboratories and the Global Navigation Satellite System (GNSS). Their purpose is to synthesize all atomic clocks in a clock group to generate a free, paper time with increased reliability and frequency stability. Both a single clock and paper time can be considered a time scale; a time scale that is not harnessed is called a free time scale. Therefore, the purpose of a time-scaling algorithm is to "generate" a new paper time scale, typically represented by the time difference between it and a specific clock. Time-scaling algorithms typically do not consider the frequency accuracy of this time scale, leaving the task of improving its frequency accuracy to the time-scale harnessing algorithm. For a timekeeping laboratory and time-frequency system (hereafter referred to as the time-frequency system), this free paper time is typically denoted as TA(k) (where k is the laboratory designator) and can be used to monitor the performance of the atomic clocks.

[0003] Existing time-scaling algorithms fall into two main categories: 1) Weighted average algorithms, primarily including the weighted average algorithm (called ALGOS) used by the International Bureau of Weights and Measures (BIPM) and the weighted average algorithm (called AT1) used by the National Institute of Standards and Technology (NIST); 2) Kalman filter algorithms, primarily including the standard Jones-Tyron Kalman filter algorithm used by NIST in the 1980s and a series of improved Kalman filter algorithms based on this, particularly the single-state variable Kalman filter time-scale generation algorithm. Because the time scales generated by the AT1 and Kalman filter algorithms are real-time, they are particularly suitable for GNSS.

[0004] Compared to other algorithms, the Kalman filter algorithm can better filter out atomic clock noise, perform clock state estimation, and detect faults, and performs better in systems that provide real-time time services, such as GNSS. However, it still has certain limitations: changes in the number of atomic clocks in a clock group require recalculation of the time scale, and jumps in the time and frequency differences of the time scale result in unstable and discontinuous time scales, which can degrade the monitoring performance of the atomic clocks or even cause them to fail. Summary of the Invention

[0005] Based on this, it is necessary to provide a method, system and device for generating an elastic time scale for changes in atomic clock groups to address the above technical problems.

[0006] A method for generating an elastic time scale for atomic clock group changes, the method comprising: Step 1, time scale parameter initialization: construct an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale with consistent generation states; Step 2, clock group operation: automatic or manual clock group operation; Step 3, clock group time difference update: update the clock group time difference based on the clock group changes after the clock group operation; Step 4, time scale model update: If the clock group has not changed, the elastic factor graph model is updated according to the current clock group time difference and clock group configuration; if the clock group has changed, the clock configuration of the backup time scale is updated according to the clock group operation result, and the time difference between the main and backup time scales is calculated and accumulated, and it is determined whether the clock group change duration meets the preset cumulative duration; if not, the elastic factor graph model is updated according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, the time difference and frequency deviation of the time scale are estimated and compensated based on the accumulated time difference results, and then the clock configuration of the main time scale is updated according to the clock group operation results, and the backup time scale is initialized on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; Step 5, time scale calculation: By solving the updated elastic factor graph model, generate the updated main time scale and backup time scale and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

[0007] In one embodiment, constructing an elastic factor graph model for time scale generation and initializing a primary time scale and a backup time scale with consistent generation states includes: The state equation and observation equation for constructing the single state variable Kalman filter time scaling algorithm are expressed as: ; in, It's the Zhong group k Time difference, superscript T represents transpose, Is the first in the Zhong group i The time difference of the atomic clocks and , is the number of atomic clocks; is the state transition matrix; is the process noise of the clock group, Is the first in the Zhong group i Process noise of atomic clocks; It's the Zhong group k The observed time difference between moments; is the observation matrix; is the observation noise, It is i The observation noise of each group of time difference is independent of each other and ; State transition matrix , process noise covariance matrix , observation matrix and the observation noise covariance matrix They are represented as follows: ; in, is the unit matrix; i Process state noise variance of an atomic clock , Indicates the i The Allan variance of an atomic clock, is the virtual Kalman sampling interval; for The noise covariance of By converting the state equation and observation equation of the single-state variable Kalman filter time scaling algorithm into state factors and observation factors in the factor graph, converting the state variables into variable nodes, and converting the covariance of the estimation error into a priori factors, a time scale generation elastic factor graph model is constructed, which can be expressed as: ; The elastic factor graph model consists of variable nodes, state factors, observation factors, and prior factors. Variable nodes correspond to state variables, state factors correspond to state transition constraints connecting state variables at adjacent moments, observation factors correspond to Kalman sensor observation constraints connecting state variables and observations, and prior factors correspond to the covariance of the estimation error in the Kalman filter. When the clock group changes, the elastic factor graph model is updated by adding or deleting factors corresponding to the atomic clocks. Setting the initial parameters of the time scale, including initializing the atomic clock group configuration, setting the initial phase and accuracy of the time scale, the initial time difference of each atomic clock, the covariance of the estimated error, and setting the cumulative duration of the clock group change; Based on the elastic factor graph model and the initial parameters of the time scale, the primary time scale and the backup time scale with consistent state are initialized.

[0008] In one embodiment, updating the clock group time difference based on the clock group change after the clock group operation includes: Based on the results of the clock group operation, the changes in the clock group are judged. If the clock group has not changed, go directly to step 4; if an atomic clock is added, add the time difference of the newly added atomic clock to the clock group time difference, and go to step 4; if an atomic clock is removed, judge whether the duration of the clock group change meets the preset cumulative duration. If not, it is considered that the removal operation is not completed, and the time difference of the atomic clock to be removed is predicted based on the state estimation result of the atomic clock, and after adding the predicted time difference to the clock group time difference, go to step 4; otherwise, it is considered that the removal operation is completed and go to step 4.

[0009] In one embodiment, the cumulative duration is determined by the accuracy of the atomic clock time difference prediction and the accuracy of the frequency deviation obtained by least squares estimation based on the time difference between the primary and backup time scales, and is expressed as: ; in, is the weighted result, is the weighted coefficient of the cumulative duration, is the weight of the atomic clock time difference prediction accuracy, The weight for the accuracy of frequency deviation estimation; 、 、 Satisfy the following constraints respectively: ; in, For cumulative duration, is a constant threshold of 1.345, is the prediction accuracy of the atomic clock time difference, yes The normalization factor of is the accuracy of the frequency deviation estimate, yes The normalization factor of When it is minimum, the corresponding The preset cumulative duration.

[0010] In one embodiment, updating the clock configuration of the backup time scale according to the clock group operation result includes: Obtain the state variables and estimation error covariance matrix of the backup time scale before the clock group changes; If it is an atomic clock i Added atomic clock based on the state estimation result prediction in the state variable i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix i The corresponding values ​​are used to construct a new initial estimation error covariance matrix.

[0011] In one embodiment, estimating and compensating for the time difference and frequency offset on a time scale based on the accumulated time difference results includes: Based on the accumulated time difference results, least squares estimation is performed to obtain the jump value of the time difference and frequency deviation between the main and backup time scales, and the jump value is compensated to the initial phase and accuracy of the time scale to achieve compensation of the time scale time difference and frequency deviation.

[0012] In one embodiment, updating the clock configuration of the main time scale according to the clock group operation result includes: Obtain the state variables and estimation error covariance matrix of the main time scale before the clock group changes; If it is an atomic clock i Increase, add Kalman filter to the state variable through Atomic clock estimated from the time difference of group observations i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix i Corresponding values ​​are constructed to obtain a new initial estimation error covariance matrix; Among them, the Kalman filter is Atomic clock estimated from the time difference of group observations i The time difference is expressed as , the iterative calculation steps are expressed as: ; in, For the Zhong Group k -1 time difference, For the Zhong Group k -1 moment k The predicted value of the time difference between the time and the time, is the state transfer matrix; for k The estimated error matrix of the time difference of the clock group at -1 hour, for kThe prediction error matrix of the time difference of the clock group; is the process noise covariance matrix; for k The Kalman filter gain at time , is the observation matrix, is the observation noise covariance matrix, It's the Zhong group k The observed time difference, is the identity matrix, the superscript T Indicates transpose.

[0013] In one embodiment, the method further includes: Step 6, atomic clock state estimation: Atomic clock state estimation is performed using a 2-state Kalman filter or a 3-state Kalman filter, and the atomic clock time difference is predicted based on the atomic clock state estimation result.

[0014] A flexible time scale generation system for atomic clock group changes, comprising: The time scale parameter initialization module is used to build an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale to generate consistent states; Clock group operation module, used for automatic or manual clock group operation; The clock group time difference update module is used to update the clock group time difference based on the clock group changes after the clock group operation; The time scale model update module is used to update the elastic factor graph model according to the current clock group time difference and clock group configuration if the clock group has not changed; if the clock group has changed, update the clock configuration of the backup time scale according to the clock group operation result, and at the same time calculate and accumulate the time difference between the main and backup time scales, and judge whether the clock group change duration meets the preset cumulative duration; if not, update the elastic factor graph model according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, first estimate and compensate the time difference and frequency deviation of the time scale based on the accumulated time difference results, then update the clock configuration of the main time scale according to the clock group operation result, and initialize the backup time scale on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; The time scale calculation module is used to generate the updated main time scale and backup time scale by solving the updated elastic factor graph model, calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for changes in the atomic clock group.

[0015] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented: Step 1, time scale parameter initialization: construct an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale with consistent generation states; Step 2, clock group operation: automatic or manual clock group operation; Step 3, clock group time difference update: update the clock group time difference based on the clock group changes after the clock group operation; Step 4, time scale model update: If the clock group has not changed, the elastic factor graph model is updated according to the current clock group time difference and clock group configuration; if the clock group has changed, the clock configuration of the backup time scale is updated according to the clock group operation result, and the time difference between the main and backup time scales is calculated and accumulated, and it is determined whether the clock group change duration meets the preset cumulative duration; if not, the elastic factor graph model is updated according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, the time difference and frequency deviation of the time scale are estimated and compensated based on the accumulated time difference results, and then the clock configuration of the main time scale is updated according to the clock group operation results, and the backup time scale is initialized on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; Step 5, time scale calculation: By solving the updated elastic factor graph model, generate the updated main time scale and backup time scale and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

[0016] The above-mentioned elastic time scale generation method, system, and device for atomic clock group changes construct an elastic factor graph model for time scale generation based on the single-state variable Kalman filter time scale algorithm. This model generates two consistent primary and backup time scales. The system then uses the accumulated time difference between the primary and backup time scales to compensate for the time difference and frequency deviation jumps in the time scales under clock group changes, thus generating an elastic time scale. This ensures the continuity and stability of the time scale and improves the monitoring performance of atomic clocks. Furthermore, the plug-and-play architecture of the factor graph allows for dynamic addition or removal of factor nodes to update the model when the clock group changes, without the need to reconstruct the entire elastic factor graph model. This makes the generation of the elastic time scale flexible and scalable. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A schematic flow chart of a method for generating an elastic time scale for atomic clock group changes according to one embodiment; Figure 2 Schematic diagram of a factor graph model of a Kalman filter in one embodiment; Figure 3 A schematic diagram of an elastic factor graph model for time scale generation in one embodiment; Figure 4 Schematic diagram of frequency stability of atomic clocks in an atomic clock group according to one embodiment; Figure 5 Schematic diagram of time scale change under clock-eliminating experiment in one embodiment; wherein, Figure 5 (a) is the overall change of time difference; Figure 5 (b) is the local change of time difference; Figure 5 (c) is the overall change in frequency deviation; Figure 5 (d) is the local change of frequency deviation; Figure 6 Schematic diagram of the time difference between the main time scale and the backup time scale in a clock-eliminating experiment in one embodiment; wherein, Figure 6 (a) is the overall time difference result; Figure 6 (b) is the local time difference result before and after the clock group changes; Figure 7 A schematic diagram of frequency stability of a reference time scale, a main time scale, and a backup time scale in a clock-eliminating experiment in one embodiment; Figure 8 Schematic diagram of time scale change under a clock experiment in one embodiment; wherein, Figure 8 (a) is the overall change of time difference; Figure 8 (b) is the local change of time difference; Figure 8 (c) is the overall change in frequency deviation; Figure 8 (d) is the local change of frequency deviation; Figure 9 Schematic diagram of the time difference results of the main time scale and the backup time scale under the clock experiment in one embodiment; wherein, Figure 9 (a) is the overall time difference result; Figure 9 (b) is the local time difference result before and after the clock group changes; Figure 10 A schematic diagram of frequency stability of a reference time scale, a main time scale, and a backup time scale in a clock-added experiment in one embodiment; Figure 11 Schematic diagram of the time scale change under the hybrid clock operation experiment; Figure 11 (a) is the overall change of time difference; Figure 11 (b) is the local change of time difference; Figure 11 (c) is the overall change in frequency deviation; Figure 11 (d) is the local change of frequency deviation; Figure 12Schematic diagram of the time difference between the main time scale and the backup time scale in the hybrid clock operation experiment; Figure 12 (a) is the overall time difference result; Figure 12 (b) is the local time difference result before and after the clock group changes; Figure 13 Schematic diagram of the frequency stability of the reference time scale, main time scale, and backup time scale in the hybrid clock operation experiment; Figure 14 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0019] In one embodiment, Figure 1 As shown, a method for generating an elastic time scale for atomic clock group changes is provided, comprising the following steps: Step 1, time scale parameter initialization: construct an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale with consistent generation states.

[0020] Step 2, clock group operation: the clock group operation is performed automatically by the time frequency system or manually by the user.

[0021] Step 3, clock group time difference update: The clock group time difference is updated based on the clock group changes after the clock group operation. Specifically, the clock group changes are judged based on the clock group operation results. If the clock group has not changed, the process proceeds directly to step 4. If an atomic clock is added, the time difference of the newly added atomic clock is added to the clock group time difference, and the process proceeds to step 4. If an atomic clock is removed, the clock group change duration is judged to see whether it meets the preset cumulative duration. If not, the removal operation is considered incomplete. The time difference of the atomic clock to be removed is predicted based on the atomic clock state estimation result, and the predicted time difference is added to the clock group time difference, and the process proceeds to step 4. Otherwise, the removal operation is considered complete and the process proceeds to step 4.

[0022] Step 4, time scale model update: If the clock group has not changed, the elastic factor graph model is updated according to the current clock group time difference and clock group configuration; if the clock group has changed, the clock configuration of the backup time scale is updated according to the clock group operation result, and the time difference between the main and backup time scales is calculated and accumulated, and it is determined whether the clock group change duration meets the preset cumulative duration; if not, the elastic factor graph model is updated according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, the time difference and frequency deviation of the time scale are first estimated and compensated based on the accumulated time difference results, and then the clock configuration of the main time scale is updated according to the clock group operation results, and the backup time scale is initialized on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model.

[0023] Step 5, time scale calculation: By solving the updated elastic factor graph model, generate the updated main time scale and backup time scale and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

[0024] In one embodiment, step 1 includes: The state equation and observation equation for constructing the single state variable Kalman filter time scaling algorithm are expressed as: ; in, It's the Zhong group k Time difference, superscript T represents transpose, Is the first in the Zhong group i The time difference of the atomic clocks and , is the number of atomic clocks; is the state transition matrix; is the process noise of the clock group, Is the first in the Zhong group i Process noise of atomic clocks; It's the Zhong group k The observed time difference between moments; is the observation matrix; is the observation noise, It is i The observation noise of each group of time difference is independent of each other and ; State transition matrix , process noise covariance matrix , observation matrix and the observation noise covariance matrix They are represented as follows: ; in, is the unit matrix; i Process state noise variance of an atomic clock , Indicates the i The Allan variance of an atomic clock, is the virtual Kalman sampling interval; for The noise covariance of

[0025] like Figure 2 As shown, there is an equivalent conversion relationship between the factor graph and the Kalman filter. The state equation and observation equation of the Kalman filter can be converted into the state factor and observation factor in the factor graph. The state variable corresponds to the variable node. The covariance of the estimated error can be converted into a priori factors, and the construction is as follows Figure 3 The elastic factor graph model generated by the time scale shown is expressed as: ; Among them, the elastic factor graph model is composed of variable nodes, state factors, observation factors and prior factors. The variable nodes correspond to state variables, the state factors correspond to the state transition constraints connecting the state variables at adjacent moments, the observation factors correspond to the Kalman sensor observation constraints connecting the state variables and the observation values, and the prior factors correspond to the covariance of the estimation error in the Kalman filter. Among them, when the clock group changes, the elastic factor graph model is updated by adding or deleting factors corresponding to the atomic clocks.

[0026] Further solvable Figure 2 The factor graph model of the Kalman filter shown and Figure 3 The elastic factor graph model shown in Figure 2. In the factor graph model of the Kalman filter, the error variable is constructed as follows: ; And the objective function of least squares is constructed as: .in, 、 represent the state error and observation error respectively.

[0027] In the elastic factor graph model, the error variable is constructed as follows: ; And the objective function of least squares is constructed as: .in, 、 Respectively represent i The state error and observation error of an atomic clock.

[0028] By solving the above two least squares objective functions, we can see that the solution results are consistent, which proves that Figure 2 The factor graph model of the Kalman filter shown and Figure 3 The elasticity factor graph models shown are equivalent.

[0029] It should be understood that a factor graph is a binary graphical model that decomposes complex global functions into products of local functions. A factor graph consists of variable nodes (random variables) and factor nodes (local functions), with edges representing dependencies between variables and factors. A factor graph can be considered a generalized form of the Kalman filter. The recursive estimation process (prediction + update) of the Kalman filter can be implemented through Gaussian message passing in the factor graph. When both process noise and observation noise are zero-mean, the state mean estimate of the factor graph is consistent with the posterior estimate of the Kalman filter. While the Kalman filter typically assumes a linear Gaussian system, the factor graph can naturally represent nonlinear, non-Gaussian systems and handle complex models through iterative optimization, improving robustness and adaptability. Factor graphs are also flexible and scalable. They utilize a plug-and-play architecture that allows factor nodes to be dynamically added or removed without reconfiguring the entire model.

[0030] Therefore, when updating the elastic factor graph model generated by the time scale based on the single-state variable Kalman filter time scale algorithm, the atomic clock information and its time difference are obtained based on the clock group time difference information and clock group configuration information input in step 3, and the different clock time differences are used to increase the Figure 3 The observation factors (time differences) and state factors of different clocks in the elastic factor graph model shown here update the prior factors in the graph. Simply put, different clock group operations result in different clock group configurations and time difference information. For model updates, only the input changes, which have no impact on the model update operation. This is a new approach for increasing the flexibility of adding or removing atomic clocks from a clock group.

[0031] Furthermore, based on the elastic factor graph model and preset time scale initial parameters, the primary and backup time scales with consistent states are initialized and generated, which can be denoted as TA(primary) and TA(backup). The time scale initial parameters include initializing the atomic clock group configuration, setting the initial phase and accuracy of the time scale, the initial time difference of each atomic clock, the covariance of the estimated errors, and setting the cumulative duration of the clock group changes.

[0032] In one embodiment, according to the method proposed in the present application, the preset cumulative duration depends on two aspects: on the one hand, the accuracy of the atomic clock time difference prediction, and on the other hand, the accuracy of the least squares estimation of the frequency deviation of the TA (master) and TA (backup) time difference series. For the atomic clock time difference prediction, its error is the deviation between the predicted atomic clock time difference and the actual measurement result, and its prediction accuracy is the root mean square (RMS) of the error sequence. For the accuracy of the frequency deviation estimation, it is the difference between the estimation result of the previous epoch and the estimation result of the next epoch. The smaller the difference, the closer it is to the theoretical true value, and the higher its accuracy. Therefore, the cumulative duration is determined by the accuracy of the atomic clock time difference prediction and the accuracy of the frequency deviation obtained by the least squares estimation based on the time difference of the master and backup time scales, expressed as: ; in, is the weighted result, is the weighted coefficient of the cumulative duration, is the weight of the atomic clock time difference prediction accuracy, is the weight of the frequency deviation estimation accuracy, each weight can be set by different users according to the importance of different parameters, and 、 、 Satisfy the following constraints respectively: ; in, For cumulative duration, is a constant threshold of 1.345, is the prediction accuracy of the atomic clock time difference, yes The normalization factor is usually 0.1ns; is the accuracy of the frequency deviation estimate, yes The normalization factor is usually 1×10 −16 s / s.

[0033] It is understandable that the weighting coefficient and cumulative duration In addition, the weighting coefficient The starting value of cannot be too small and cannot differ from other values ​​by an order of magnitude. Therefore, this application modifies the standard S curve to obtain the cumulative time The weighting coefficient . and It is modified based on the inverse function of the Huber weight function. and It is set by the user based on the time-frequency system status and empirical statistical data. In short, the R value is based on different cumulative time Calculated, when R is the smallest, its corresponding It is the most reasonable value calculated, which is the preset cumulative duration.

[0034] In one embodiment, updating the clock configuration of the backup time scale according to the clock group operation result includes: Obtain the state variables and estimation error covariance matrix of the backup time scale before the clock group changes; If it is an atomic clock i Added atomic clock based on the state estimation result prediction in the state variable i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix i The corresponding values ​​are used to construct a new initial estimation error covariance matrix.

[0035] In one embodiment, estimating and compensating for the time difference and frequency offset on a time scale based on the accumulated time difference results includes: Based on the accumulated time difference results, least squares estimation is performed to obtain the jump value of the time difference and frequency deviation between the main and backup time scales, and the jump value is compensated to the initial phase and accuracy of the time scale to achieve compensation of the time scale time difference and frequency deviation.

[0036] In one embodiment, updating the clock configuration of the main time scale according to the clock group operation result includes: Obtain the state variables and estimation error covariance matrix of the main time scale before the clock group changes; If it is an atomic clock i Increase, add Kalman filter to the state variable through Atomic clock estimated from the time difference of group observations i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix iCorresponding values ​​are constructed to obtain a new initial estimation error covariance matrix; Among them, the Kalman filter is Atomic clock estimated from the time difference of group observations i The time difference is expressed as , the iterative calculation steps are expressed as: ; in, For the Zhong Group k -1 time difference, For the Zhong Group k -1 moment k The predicted value of the time difference between the time and the time, is the state transfer matrix; for k The estimated error matrix of the time difference of the clock group at -1 hour, for k The prediction error matrix of the time difference of the clock group; is the process noise covariance matrix; for k The Kalman filter gain at time , is the observation matrix, is the observation noise covariance matrix, It's the Zhong group k The observed time difference, is the identity matrix, the superscript T Indicates transpose.

[0037] Furthermore, after updating the clock configuration of the primary time scale, the calculation results of the primary time scale after the clock configuration update can be used to update the parameters of the backup time scale and initialize it to ensure the consistency of TA (primary) and TA (backup).

[0038] In one embodiment, the method further includes: Step 6: Atomic clock state estimation. It should be understood that using a Kalman filter to estimate the atomic clock state and predict the time difference is a mature algorithm. Typically, a two-state Kalman filter or a three-state Kalman filter is used to estimate the atomic clock state and predict the atomic clock time difference based on the atomic clock state estimation result.

[0039] Among them, the state transfer matrix of the 2-state Kalman filter is , process noise covariance matrix , observation matrix and the observation noise covariance matrix They are represented as follows: .

[0040] State transition matrix of 3-state Kalman filter , process noise covariance matrix , observation matrix and the observation noise covariance matrix They are represented as follows: ; in, is the time sampling interval; For atomic clocks i The standard deviation of the observation noise. 、 and is the noise parameter, and the Allan variance inversion method is generally used to obtain the specific value of the noise parameter.

[0041] In summary, the above-mentioned elastic time scale generation method for atomic clock group changes builds an elastic factor graph model for time scale generation based on the single-state variable Kalman filter time scale algorithm. This generates two consistent time scales, primary and backup, and uses the accumulated time difference between the primary and backup time scales under atomic clock group changes to compensate for the time difference and frequency deviation jumps in the time scale under clock group changes. This generates an elastic time scale, ensuring the continuity and stability of the time scale and improving the monitoring performance of atomic clocks. Furthermore, by leveraging the plug-and-play architecture of the factor graph, factor nodes can be dynamically added or removed to update the model when the clock group changes, without having to reconstruct the entire elastic factor graph model. This makes the generation of the elastic time scale flexible and scalable.

[0042] The performance of the method proposed in this application was further verified experimentally. Specifically, relevant experimental data were obtained from a stably operating time and frequency system. The atomic clock group includes 6 hydrogen clocks, denoted as H01 to H06; and 3 cesium clocks, denoted as Cs01 to Cs03. The time difference of each atomic clock in the atomic clock group is measured using a high-precision time interval measurement device with a sampling interval of 1 s and a data length of 45 days. The time difference between the real-time physical signal and the coordinated universal time UTC (k) is obtained through two-way time transmission of optical fiber. Using the triangular hat method, the frequency stability performance of each clock is obtained as follows: Figure 4 shown. Figure 4 Among them, the performance of H01~H03 and H06 is comparable, the performance of H04 and H05 is comparable, slightly worse than other hydrogen clocks; the performance of Cs01~Cs03 is comparable.

[0043] To demonstrate the effectiveness of the proposed method, six of the nine atomic clocks in the group, H01 to H04, Cs01, and Cs02, were selected in subsequent experiments to construct an original atomic clock group. The experiments shown in Table 1 were conducted on this group configuration. The obtained time difference with UTC(k) was also used to evaluate the frequency stability of the generated time scale.

[0044] Table 1 Experimental settings

[0045] 1. Cumulative duration selection experiment: Under the parameter settings of the typical time scale calculation interval of the time-frequency system is 300 s and the virtual sampling interval is 86400 s, different cumulative durations within 24 hours are calculated. t The results are shown in Table 2. It can be seen that as the accumulation time increases, the prediction accuracy of the time difference decreases, while the frequency deviations estimated by the TA (primary) and TA (backup) decrease. Furthermore, the accumulation time when R is minimum is 7 hours. Therefore, the accumulation time selected for subsequent experiments is 7 hours.

[0046] Table 2 Cumulative duration calculation results

[0047] In subsequent experiments, the experimental parameters are shown in Table 3. The cumulative duration in the method proposed in this application is determined by the cumulative duration selection experiment. The virtual sampling interval and time scale calculation interval are set according to the construction indicators and actual operation conditions of the time-frequency system.

[0048] Table 3 Experimental parameters

[0049] 2. Clock Elimination Experiment: In this experiment, we construct scenarios where the six clock groups are eliminated, based on the atomic clock types and their weights, as shown in Table 4. The time scale without clock group changes is used as the reference time scale, denoted as TA(reference).

[0050] Table 4 Scenario settings for the clock elimination experiment

[0051] like Figure 5 As shown in , before the clock group changes, the generated elastic time scale is consistent with the reference time scale, and after the clock group changes, the time difference changes. Figure 5 (a) and Figure 5 As shown in (c), the trends of scenes Del 1 and Del 5 are basically the same; scene Del 4 has a part that is consistent with the trends of scenes Del 2 and Del 6, but then separates. The frequency deviation after the clock group changes in all scenes is better than 1×10 −15 s / s. Figure 5 (b) and Figure 5 As shown in (d), the time difference jump in scenario Del 3 can be ignored; the trends in scenarios Del 2 and Del 6 are basically the same, with a small time difference jump; the trends in scenarios Del 2, Del 4, and Del 6 are the same, with a large time difference jump; the time difference jump values ​​before and after the clock group change are all better than 20 ps, ​​and the jump value of the frequency deviation is better than 1×10 −17 s / s. According to the experiment set up in Table 4, scenario Del 3 only eliminated Cs02, scenarios Del 2 and Del 6 eliminated H04, scenarios Del 1 and Del 5 eliminated H03, and scenario Del 4 eliminated both H03 and H04. Furthermore, based on theoretical and experimental data analysis and the set experimental parameters, the weight of hydrogen clocks in the generated time scale is significantly higher than that of cesium clocks, with H03 having a higher weight than H04. The weight of hydrogen clocks in the generated time scale is significantly higher than that of cesium clocks, with H03 having a higher weight than H04. Therefore, when atomic clocks with lower weights are eliminated from the clock group, the time scale changes slightly; when atomic clocks with higher weights are eliminated from the clock group, the time scale changes significantly. When multiple clocks are eliminated, the magnitude of the change in time difference depends on the weight ratio of the atomic clocks, that is, their performance.

[0052] The time difference results of TA (master) and TA (backup) under different clock elimination test scenarios are as follows: Figure 6 As shown. Figure 6 As shown in the figure, before the clock group change, TA (master) and TA (backup) were aligned. After the clock group change, their time difference changed. After the clock group change, the time difference between TA (master) and TA (backup) changed by less than 0.4ns. When the cumulative time is met, the time difference between TA (master) and TA (backup) is quickly adjusted and maintained at a level better than 0.1ns.

[0053] The frequency stability results of TA (reference), TA (master) and TA (backup) under the clock elimination experiment are as follows: Figure 7 shown. Figure 7 In the figure, Ref, Master, and Backup represent TA (reference), TA (master), and TA (backup), respectively. Their frequency stability is basically the same. Although TA (backup) undergoes time difference adjustment during the clock group change process, the adjustment amount is less than 0.3ns and the adjustment time is relatively short (7 hours). Therefore, the impact on long-term data (45 days) is relatively small. Therefore, the stability performance is as follows: Figure 7 shown.

[0054] Table 5 Adding the bell experiment scene settings

[0055] 3. Clock Addition Experiment: In this experiment, we construct scenarios where clocks are added to six clock groups, as shown in Table 5, based on the types of atomic clocks and their weights. The time scale without clock group changes is used as the reference time scale, denoted as TA(reference).

[0056] For the above experimental scenario, the method proposed in this application is used to generate the elastic time scale. The time scale changes generated under different clock addition experimental scenarios are as follows: Figure 8 As shown. Figure 8 As shown in , before the clock group changes, the generated elastic time scale is consistent with the reference time scale, and after the clock group changes, its time difference and frequency deviation change. Figure 8 (a) and Figure 8 As shown in (c), the trends of scenarios Add 2 and Add 6 are basically the same; the trends of scenarios Add 1 and Add 5 are basically the same; scenario Add 4 has a portion that is consistent with the trends of scenarios Add 2 and Add 6, but then separates. The time scale change is the largest in scenario Add 4. The frequency deviation after the clock group change in all scenarios is better than 3×10 −16 s / s. Figure 8 (b) and Figure 8 As shown in (d), the time difference jump value after the clock group changes in all scenarios is better than 10ps, and the frequency deviation jump value is better than 4×10 −18 s / s. According to the experimental setup in Table 5, scenario Add 3 only adds CsO2, scenarios Add 2 and Add 6 both add HO6, scenarios Add 1 and Add 5 both add HO5, and scenario Add 4 adds both H05 and H06. Furthermore, based on theoretical and experimental data analysis and the set experimental parameters, the weight of hydrogen clocks in the generated time scale is significantly higher than that of cesium clocks, and H06 has a higher weight than H05. Therefore, when a clock group adds an atomic clock with a lower weight, the time scale difference changes slightly; when a clock group adds an atomic clock with a higher weight, the time scale changes significantly. When multiple clocks are added, the extent of the time scale change depends on the weight ratio of the atomic clocks, that is, their performance.

[0057] The time difference between TA (master) and TA (backup) under the added clock experiment is as follows: Figure 9 shown. Figure 9 Before the clock group changes, the time difference between the primary and backup devices coincides. After the clock group changes, the time difference between the primary and backup devices changes. The trend of change varies in each scenario, but the change is less than 0.3ns. Once the cumulative duration is met, the time difference between the primary and backup devices adjusts rapidly and remains below 0.1ns.

[0058] The frequency stability results of TA (reference), TA (master) and TA (backup) under the added clock experiment are as follows: Figure 10 shown. Figure 10 In the figure, Ref, Master, and Backup represent TA (reference), TA (master), and TA (backup), respectively. Their frequency stability is basically the same. Although TA (backup) undergoes time difference adjustment during the clock group change process, the adjustment amount is less than 0.3ns and the adjustment time is relatively short (7 hours). Therefore, the impact on long-term data (45 days) is relatively small, so the stability is as follows: Figure 10 shown.

[0059] 4. Hybrid clock operation experiment: In this experiment, we construct a hybrid operation scenario of adding / removing clocks from the six clock groups shown in Table 6, based on the atomic clock types and their weights. The time scale without clock group changes is used as the reference time scale, denoted as TA(reference).

[0060] Table 6 Hybrid clock operation experiment scenario settings

[0061] For the above experimental scenario, the method proposed in this application is used to generate the elastic time scale. The time scale changes generated in the experimental scenario of different hybrid clock operations are as follows: Figure 11 shown. Figure 11 In the example, before the clock group changes, the generated elastic time scale is consistent with the reference time scale. After the clock group changes, the time difference and frequency deviation change. Figure 11 (a) and Figure 11 As shown in (c), the trends of scenarios Mix 1 and Mix 5 are basically the same; scenario Mix 4 has a portion that is consistent with the trends of scenarios Mix 1 and Mix 5, but then separates. Scenario Mix 6 has the largest changes in time difference and frequency deviation. The frequency deviation after the clock group change in all scenarios is better than 1.5×10 −15 s / s. Figure 11 (b) and Figure 11 As shown in (d), the time difference jump value after the clock group changes in all scenarios is better than 20ps, and the frequency deviation jump value is better than 1×10 −17s / s. According to the experiment set up in Table 6, scenario Mix 1 removes and adds a hydrogen clock of comparable performance, while scenario Mix 2 removes and adds a cesium clock of comparable performance. Scenarios Mix 1, Mix 3, and Mix 4 all remove clock H04 and add atomic clocks of varying weights. Scenarios Mix 5 and Mix 6 increase the number of atomic clocks removed and added. Furthermore, based on theoretical and experimental data analysis and the set experimental parameters, the weight of hydrogen clocks in the generated time scale is significantly higher than that of cesium clocks, with H03 and H06 having higher weights than H04 and H05. Therefore, when adding or removing atomic clocks with lower weights from the clock group, the time scale difference changes slightly; when adding or removing atomic clocks with higher weights, the time scale difference changes significantly. When multiple clocks are added or removed, the magnitude of the time scale change depends on the weight ratio of the atomic clocks, that is, their performance.

[0062] The time difference results of TA (master) and TA (backup) under different hybrid clock operation experimental scenarios are as follows: Figure 12 shown. Figure 12 Before the clock group changes, the time difference between the primary and backup devices coincides. After the clock group changes, the time difference between the primary and backup devices changes. The trend of change varies in each scenario, but the change is less than 0.3ns. Once the cumulative time is met, the time difference between the primary and backup devices adjusts rapidly and remains below 0.1ns.

[0063] The frequency stability results of TA (reference), TA (master) and TA (backup) in the hybrid clock operation experiment are as follows: Figure 13 shown. Figure 13 In the figure, Ref, Master, and Backup represent TA (reference), TA (master), and TA (backup), respectively. Their frequency stability is basically the same. Although TA (backup) undergoes time difference adjustment during the clock group change process, the adjustment amount is less than 0.3ns and the adjustment time is relatively short (7 hours). Therefore, the impact on long-term data (45 days) is relatively small, so the stability is as follows: Figure 13 shown.

[0064] In summary, the elastic time scale generated by this application changes the time difference and frequency deviation under the clock group changes, which include the removal and addition of atomic clocks. The performance of the removed and added atomic clocks is proportional to the impact on the time scale: the better the atomic clock performance, the greater the time scale change, and the worse the performance, the smaller the time scale change. In all experimental scenarios, the jump value of the elastic time scale time difference is better than 20 ps, ​​and the jump value of the frequency deviation is better than 1×10 −17 s / s, frequency deviation is better than 1.5×10 −15s / s. Meanwhile, the time difference between the TA (backup) and TA (master) generated using the elastic scale generation method is better than 0.3ns. After the clock group change operation is completed, the time difference between the generated TA (master) and TA (backup) remains better than 0.1ns, and their frequency stability remains consistent with no degradation.

[0065] In one embodiment, a system for generating a flexible time scale responsive to changes in an atomic clock group is provided, comprising: The time scale parameter initialization module is used to build an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale to generate consistent states; Clock group operation module, used for automatic or manual clock group operation; The clock group time difference update module is used to update the clock group time difference based on the clock group changes after the clock group operation; The time scale model update module is used to update the elastic factor graph model according to the current clock group time difference and clock group configuration if the clock group has not changed; if the clock group has changed, update the clock configuration of the backup time scale according to the clock group operation result, and at the same time calculate and accumulate the time difference between the main and backup time scales, and judge whether the clock group change duration meets the preset cumulative duration; if not, update the elastic factor graph model according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, first estimate and compensate the time difference and frequency deviation of the time scale based on the accumulated time difference results, then update the clock configuration of the main time scale according to the clock group operation result, and initialize the backup time scale on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; The time scale calculation module is used to generate the updated main time scale and backup time scale by solving the updated elastic factor graph model, calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for changes in the atomic clock group.

[0066] For the specific limitations of the elastic time scale generation system for atomic clock group changes, please refer to the limitations of the elastic time scale generation method for atomic clock group changes above, which will not be repeated here. Each module in the above-mentioned elastic time scale generation system for atomic clock group changes can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0067] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 14 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for generating an elastic time scale for changes in an atomic clock group is implemented. The display screen of the computer device can be a liquid crystal display or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad provided on the computer device housing, or an external keyboard, touchpad or mouse.

[0068] Those skilled in the art will understand that Figure 14 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0069] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following steps are implemented: Step 1, time scale parameter initialization: construct an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale with consistent generation states; Step 2, clock group operation: automatic or manual clock group operation; Step 3, clock group time difference update: update the clock group time difference based on the clock group changes after the clock group operation; Step 4, time scale model update: If the clock group has not changed, the elastic factor graph model is updated according to the current clock group time difference and clock group configuration; if the clock group has changed, the clock configuration of the backup time scale is updated according to the clock group operation result, and the time difference between the main and backup time scales is calculated and accumulated, and it is determined whether the clock group change duration meets the preset cumulative duration; if not, the elastic factor graph model is updated according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, the time difference and frequency deviation of the time scale are estimated and compensated based on the accumulated time difference results, and then the clock configuration of the main time scale is updated according to the clock group operation results, and the backup time scale is initialized on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; Step 5, time scale calculation: By solving the updated elastic factor graph model, generate the updated main time scale and backup time scale and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

[0070] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0071] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The above embodiments merely illustrate several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application.

Claims

1. A method for generating an elastic time scale for atomic clock group changes, characterized in that: The method comprises: Step 1, time scale parameter initialization: construct an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale with consistent generation states; Step 2, automatically or manually operate the clock group; Step 3: Update the time difference of the clock group based on the changes of the clock group after the clock group operation; Step 4, time scale model update: if the clock group has not changed, the elastic factor graph model is updated according to the current clock group time difference and clock group configuration; if the clock group has changed, the clock configuration of the backup time scale is updated according to the clock group operation result, and the time difference between the main and backup time scales is calculated and accumulated, and it is determined whether the clock group change duration meets the preset cumulative duration; if not, the elastic factor graph model is updated according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, the time difference and frequency deviation of the time scale are estimated and compensated based on the accumulated time difference results, and then the clock configuration of the main time scale is updated according to the clock group operation results, and the backup time scale is initialized on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; Step 5, time scale calculation: By solving the updated elastic factor graph model, generate the updated main time scale and backup time scale and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

2. The method for generating an elastic time scale for atomic clock group changes according to claim 1, characterized in that: Construct an elastic factor graph model for time scale generation and initialize the primary and backup time scales with consistent generation states, including: The state equation and observation equation for constructing the single state variable Kalman filter time scaling algorithm are expressed as: ； in, It's the Zhong group k Time difference, superscript T represents transpose, Is the first in the Zhong group i The time difference of the atomic clocks and , is the number of atomic clocks; is the state transition matrix; is the process noise of the clock group, Is the first in the Zhong group i Process noise of atomic clocks; It's the Zhong group k The observed time difference between moments; is the observation matrix; is the observation noise, It is i The observation noise of each group of time difference is independent of each other and ; State transition matrix , process noise covariance matrix , observation matrix and the observation noise covariance matrix They are represented as follows: ； in, is the unit matrix; i Process state noise variance of an atomic clock , Indicates the i The Allan variance of an atomic clock, is the virtual Kalman sampling interval; for The noise covariance of By converting the state equation and observation equation of the single-state variable Kalman filter time scaling algorithm into state factors and observation factors in the factor graph, converting the state variables into variable nodes, and converting the covariance of the estimation error into a priori factors, a time scale generation elastic factor graph model is constructed, which can be expressed as: ； The elastic factor graph model is composed of variable nodes, state factors, observation factors, and prior factors. The variable nodes correspond to state variables, the state factors correspond to state transition constraints connecting state variables at adjacent moments, the observation factors correspond to Kalman sensor observation constraints connecting state variables and observation values, and the prior factors correspond to the covariance of the estimation error in the Kalman filter. When the clock group changes, the elastic factor graph model is updated by adding or deleting factors corresponding to the atomic clocks. Setting the initial parameters of the time scale, including initializing the atomic clock group configuration, setting the initial phase and accuracy of the time scale, the initial time difference of each atomic clock, the covariance of the estimated error, and setting the cumulative duration of the clock group change; Based on the elastic factor graph model and the initial parameters of the time scale, a primary time scale and a backup time scale with consistent state are initialized and generated.

3. The method for generating an elastic time scale for atomic clock group changes according to claim 1, characterized in that: Update the clock group time difference based on the clock group changes after the clock group operation, including: Based on the results of the clock group operation, the changes in the clock group are judged. If the clock group has not changed, go directly to step 4; if an atomic clock is added, add the time difference of the newly added atomic clock to the clock group time difference, and go to step 4; if an atomic clock is removed, judge whether the duration of the clock group change meets the preset cumulative duration. If not, it is considered that the removal operation is not completed, and the time difference of the atomic clock to be removed is predicted based on the state estimation result of the atomic clock, and after adding the predicted time difference to the clock group time difference, go to step 4; otherwise, it is considered that the removal operation is completed and go to step 4.

4. The method for generating an elastic time scale for atomic clock group changes according to claim 3, characterized in that: The cumulative duration is determined by the accuracy of the atomic clock time difference prediction and the accuracy of the frequency deviation obtained by the least squares estimation based on the time difference between the primary and backup time scales, and is expressed as: ； in, is the weighted result, is the weighted coefficient of the cumulative duration, is the weight of the atomic clock time difference prediction accuracy, The weight for the accuracy of frequency deviation estimation; 、 、 Satisfy the following constraints respectively: ； in, For cumulative duration, is a constant threshold of 1.345, is the prediction accuracy of the atomic clock time difference, yes The normalization factor of is the accuracy of the frequency deviation estimate, yes The normalization factor of When it is minimum, the corresponding The preset cumulative duration.

5. The method for generating an elastic time scale for atomic clock group changes according to claim 1, characterized in that: Update the clock configuration of the standby time scale according to the clock group operation results, including: Obtain the state variables and estimation error covariance matrix of the backup time scale before the clock group changes; If it is an atomic clock i Added atomic clock based on the state estimation result prediction in the state variable i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix i The corresponding values ​​are used to construct a new initial estimation error covariance matrix.

6. The method for generating an elastic time scale for atomic clock group changes according to claim 1, characterized in that: Estimation and compensation of time scale time difference and frequency deviation based on the accumulated time difference results, including: Based on the accumulated time difference results, least squares estimation is performed to obtain the jump value of the time difference and frequency deviation between the main and backup time scales, and the jump value is compensated to the initial phase and accuracy of the time scale to achieve compensation of the time scale time difference and frequency deviation.

7. The method for generating an elastic time scale for atomic clock group changes according to claim 1, characterized in that: Update the clock configuration of the main time scale according to the results of the clock group operation, including: Obtain the state variables and estimation error covariance matrix of the main time scale before the clock group changes; If it is an atomic clock i Increase, add Kalman filter to the state variable through Atomic clock estimated from the time difference of group observations i The time difference is used to construct a new initial state variable, and the mean of the diagonal elements of the estimated error covariance matrix is ​​added to the estimated error covariance matrix as the new initial estimated error covariance matrix; If it is an atomic clock i Eliminate, eliminate the atomic clock in the state variables i The value of , constructs the new initial state variables, and removes the atomic clock in the estimated error covariance matrix i Corresponding values ​​are constructed to obtain a new initial estimation error covariance matrix; Among them, the Kalman filter is Atomic clock estimated from the time difference of group observations i The time difference is expressed as , the iterative calculation steps are expressed as: ； in, For the Zhong Group k -1 time difference, For the Zhong Group k -1 moment k The predicted value of the time difference between the time and the time, is the state transfer matrix; for k The estimated error matrix of the time difference of the clock group at -1 hour, for k The prediction error matrix of the time difference of the clock group; is the process noise covariance matrix; for k The Kalman filter gain at time , is the observation matrix, is the observation noise covariance matrix, It's the Zhong group k The observed time difference, is the identity matrix, the superscript T Indicates transpose.

8. The method for generating an elastic time scale for atomic clock group changes according to claim 1 or 5, characterized in that: The method further comprises: Step 6, atomic clock state estimation: Atomic clock state estimation is performed using a 2-state Kalman filter or a 3-state Kalman filter, and the atomic clock time difference is predicted based on the atomic clock state estimation result.

9. A flexible time scale generation system for atomic clock group changes, characterized by: The system comprises: The time scale parameter initialization module is used to build an elastic factor graph model for time scale generation and initialize the primary time scale and backup time scale to generate consistent states; Clock group operation module, used for automatic or manual clock group operation; The clock group time difference update module is used to update the clock group time difference based on the clock group changes after the clock group operation; The time scale model update module is used to update the elastic factor graph model according to the current clock group time difference and clock group configuration if the clock group has not changed; if the clock group has changed, update the clock configuration of the backup time scale according to the clock group operation result, and at the same time calculate and accumulate the time difference between the main and backup time scales, and judge whether the clock group change duration meets the preset cumulative duration; if not, update the elastic factor graph model according to the current time difference between the main and backup time scales and the clock group configuration; if satisfied, first estimate and compensate the time difference and frequency deviation of the time scale based on the accumulated time difference result, then update the clock configuration of the main time scale according to the clock group operation result, and initialize the backup time scale on the main time scale after the clock configuration is updated. After maintaining the consistency of the main and backup time scales, the optimized time difference result between the main and backup time scales is obtained and combined with the current clock group configuration to update the elastic factor graph model; The time scale calculation module is used to generate the updated main time scale and backup time scale by solving the updated elastic factor graph model and calculate the time difference between the updated main and backup time scales, and output the updated main time scale as the elastic time scale for the changes of the atomic clock group.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

Images