On-satellite time synchronization method and system based on pulse per second

Abstract

The invention belongs to the technical field of spacecraft electronics and control systems, and discloses an on-satellite time synchronization method and system based on pulse per second. The method comprises the following steps: S1, collecting N + 1 continuous external standard pulses per second, and recording a count value corresponding to a local clock of an on-satellite computer when each pulse per second arrives to form a timestamp sequence; s2, calculating actual cumulative time of a local clock in N standard pulse per second intervals based on the timestamp sequence; calculating a time scaling coefficient based on the standard time interval of the N standard pulses per second and the actual cumulative time of the local clock in the N standard pulses per second intervals; and S3, taking the time scaling factor as a multiplication factor, and continuously and uniformly scaling the timing rate of the local clock of the satellite computer to obtain the corrected time. According to the method, the high-quality on-satellite time which is free of jump, accurate for a long time and uniform for a short time can be generated by utilizing a simple algorithm under the condition of not introducing extra hardware.

Description

Technical Field

[0001] This invention belongs to the field of spacecraft electronics and control systems technology, and specifically relates to an on-board time synchronization method based on second pulses. Background Technology

[0002] In spacecraft such as satellites and space stations, precise time synchronization is fundamental to ensuring the coordinated operation of various subsystems. Onboard computers typically rely on their own crystal oscillators to generate clock signals, but crystal oscillators have initial frequency deviations and frequency drifts caused by temperature and radiation, resulting in cumulative errors between the local clock and standard time (such as UTC).

[0003] In existing technologies, the main methods for resolving errors between local clocks and standard time and achieving time synchronization include the following: (1) Direct Time Reset Method. For example, patent application number CN202010546620.0 discloses a method for synchronizing the clocks of an AIRT-ROS real-time system and a non-real-time system. When the onboard computer receives a pulse-per-second (1PPS) signal from a high-precision source (such as a GPS receiver), it directly sets or clears the "second" part of the software clock to align it with the standard time. However, this method can cause sudden time jumps (backtracking or jumping). Such jumps are extremely dangerous for GNC algorithms that use time as the independent variable, and may lead to abnormal calculation of the differential terms of control commands, causing instantaneous disturbances or instability in the system. The root cause is that this method only corrects the "position" of the time point, but does not correct the "speed" (frequency) of the local clock. Therefore, between the two corrections, the clock still drifts at the wrong speed.

[0004] (2) Frequency Adjustment Method. Based on single-pulse measurement, this method is susceptible to signal jitter and instantaneous errors, resulting in poor short-term stability and difficulty in providing uniform and accurate timestamps within the millisecond-level period required by the GNC algorithm. For example, patent application number CN201910160892.4 discloses a clock synchronization method based on second pulses from a GNSS receiver. This method calculates the deviation between the instantaneous frequency of the local clock and the standard frequency by measuring the local clock count between two adjacent second pulses, and then dynamically adjusts the increment of the clock counter (i.e., the time represented by each "tick"). For example, if the local clock count is measured to be 999 values ​​between two adjacent 1PPS (the standard should be 1000), it is considered that the local clock is too slow, and the count increment will be appropriately increased in the next second. A complete hardware circuit system is required to complete this. Furthermore, the correction accuracy is greatly affected by the measurement error of a single pulse interval. The second pulse signal itself may have slight jitter, and there are uncertainties such as operating system interrupt delays when the computer records the timestamp. These noises are directly incorporated into frequency calculations, causing large fluctuations in the time scaling factor. This results in a continuous but uneven output time, exhibiting insufficient stability on the second and millisecond scales. The root cause lies in making decisions based solely on two samples, failing to statistically smooth out random errors. This prevents it from providing high-quality service to GNC systems that require highly uniform timestamps.

[0005] (3) Hardware synchronization method: By conditioning the hardware signal, such as performing a logic inversion operation on the hardware clock signal to output a clock signal with periodic phase transitions, etc. This method is not suitable for GNC algorithms that require high continuity timestamps, and requires additional hardware circuits, increasing system cost, weight and power consumption. In the on-board environment, hardware reliability may be affected by factors such as radiation and temperature, making maintenance complex.

[0006] (4) Software filtering method: Although using filtering algorithms (such as Kalman filtering) can improve accuracy, the computational complexity is high and it consumes a lot of computing resources, making it difficult to achieve real-time performance in resource-constrained on-board computers.

[0007] Existing methods all attempt to directly measure and adjust the instantaneous frequency difference between adjacent pulses, or reset the absolute time point. These methods inevitably cause time jumps, failing to meet the high precision and stability requirements of critical subsystems such as guidance, navigation, and control (GNC) for time references. In stark contrast, this invention proposes an on-board time synchronization method based on long-term statistical rate scaling. This method ensures time continuity while overcoming the shortcomings of poor short-term stability. Without introducing additional hardware, it utilizes a simple algorithm to generate high-quality on-board time that is both jump-free, long-term accurate, and short-term uniform. Summary of the Invention

[0008] To address the above problems, one objective of this invention is to provide an on-board time synchronization method based on second pulses. This method utilizes an external second pulse signal and a software algorithm to continuously and uniformly calibrate the local clock of the on-board computer. The specific technical solution of this invention is as follows: An on-board time synchronization method based on second pulses includes the following steps: S1. Signal Acquisition and Recording: Acquire N+1 consecutive external standard second pulses and record the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1; S2. Coefficient Calculation: Based on the timestamp sequence, calculate the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals; based on the standard time interval T_idea1 of the N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals, calculate the time scaling factor K; where K = T_ideal / T_local; S3. Rate scaling correction: The time scaling factor is used as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time.

[0009] Further, in step S2, the actual cumulative time T_local of the local clock within N standard second pulse intervals is calculated using the following formula: T_local=C N -C0; Among them, C0, C N These are the first and last timestamps of the timestamp sequence, respectively. Furthermore, the actual cumulative time T_local of the local clock over N standard second pulse intervals is calculated using the following method: The actual cumulative time T_local is obtained by weighted averaging of multiple consecutive timestamp intervals in the timestamp sequence.

[0010] Furthermore, in step S3, the timing rate of the onboard computer's local clock is continuously and uniformly scaled using the following formula to generate the corrected time t_corrected: t_corrected=(C_current-C_base)×K+T_base; Where C_base is the count value of the onboard computer's local clock when the standard second pulse arrives at the calibration reference point, T_base is the standard time of the corresponding standard second pulse, C_current is the count value of the current local clock to be calibrated, and K is the time scaling factor.

[0011] Furthermore, the value of N is dynamically adjusted according to the spacecraft system's requirements for accuracy and response speed; when the clock is unstable, the value of N is greater than the value of N when it is stable.

[0012] Furthermore, the value of N is 10.

[0013] Furthermore, the method further includes the step of: The time scaling factor is updated using a sliding window method: each time a new second pulse is received, the earliest one is discarded, and the time scaling factor K is recalculated based on the updated timestamp sequence.

[0014] Furthermore, the calculated new time scaling factor K_new is used after being subjected to a first-order low-pass filter with the time scaling factor K_old of the previous cycle.

[0015] Another object of the present invention is to provide an on-board time synchronization system based on second pulses, the system comprising: The signal acquisition module acquires N+1 consecutive external standard second pulses and records the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1. The caching module is used to store timestamp sequences. If the timestamp sequence is not continuous, recording will start again. The coefficient calculation module calculates the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals based on the timestamp sequence; and calculates the time scaling factor K based on the standard time interval T_idea1 of the N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals; where K=T_ideal / T_local; The rate scaling correction module uses the time scaling coefficient as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time.

[0016] Furthermore, the system also includes: The sliding update module uses a sliding window approach to discard the oldest second pulse each time a new second pulse is received. The trigger coefficient calculation module recalculates the time scaling factor K based on the updated timestamp sequence.

[0017] This invention is based on the technical concept of calculating a global, long-term average time scaling factor K based on multiple continuous standard second pulse signals, and using this factor to perform continuous and uniform scaling correction on the local clock. Compared with the prior art, one or more of the above technical solutions can achieve at least one of the following beneficial effects: (1) High short-term stability and no time jump: In view of the technical problem that the instantaneous frequency adjustment method in the existing technology corrects the instantaneous error between two adjacent pulses, the correction amount AT is discrete and abrupt, which leads to a step in the output time at the micro level, the present invention calculates a global time scaling factor K based on multiple second pulses, which effectively filters signal jitter and single measurement error, so that the output time has excellent uniformity within seconds and at the millisecond scale, perfectly meeting the requirements of the GNC algorithm; Compared with the direct time synchronization method, this invention adopts the scaling principle. Its correction process is to scale the local timeline as a whole and continuously according to the time scaling factor K, just like stretching or compressing a time axis with elastic deviation to a standard length. This eliminates micro-steps and achieves sub-second-level high uniformity. It ensures the continuity and monotonically increasing nature of the time axis and completely avoids the control system risks caused by time reset. (2) Simple and easy to implement: This method is implemented purely in software, requiring no additional hardware, thus reducing system cost and complexity. The algorithm has low computational load and is suitable for resource-constrained onboard computers; (3) Strong robustness: By averaging multiple data acquisitions, the impact of noise and outliers is reduced, and reliable synchronization performance can still be maintained under harsh on-board environments (such as temperature changes and radiation). (4) Good real-time performance: The calibration process is carried out continuously and can adapt to clock drift in real time, ensuring long-term synchronization effect.

[0018] This invention overcomes the shortcomings of the prior art and provides an efficient, reliable and low-cost on-board time synchronization solution, which is particularly suitable for spacecraft systems with high time accuracy requirements. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the on-board time synchronization method based on second pulses according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram of the time synchronization logic of the on-board time synchronization method in an embodiment of this application.

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Detailed Implementation

[0023] like Figure 1 As shown in the figure, this application provides an on-board time synchronization method based on second pulses, including the following steps: S1. Signal Acquisition and Recording: Acquire N+1 consecutive external standard second pulses and record the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1.

[0024] Specifically, the onboard computer continuously monitors the externally input high-precision pulse-per-second (1PPS) signal. Whenever the rising edge of a pulse-per-second arrives, it immediately captures and records the local clock count value C of the onboard computer at that moment. N (Where N=0, 1, 2, 3, ...), forming a timestamp sequence; or, whenever the falling edge of a second pulse arrives, immediately capture and record the local clock count value of the onboard computer at that moment, forming a timestamp sequence; until the timestamps corresponding to N+1 consecutive second pulses are buffered, the timestamp sequence is (C0, C1, C2, ..., C...). N Where N is an integer greater than 1, and the N+1 second pulses must be consecutive; if they are not consecutive, recording must start again.

[0025] In this embodiment, N is preferably 10 to achieve a balance between response speed and statistical accuracy.

[0026] In some embodiments, the value of N may not be fixed at 10, but may be dynamically adjusted according to the system's current requirements for accuracy and response speed. For example, when the system is first powered on or when the clock is considered unstable, a larger N (such as 20) is used to achieve fast convergence; in a stable state, a smaller N (such as 5) is used to track drift more quickly.

[0027] S2. Coefficient Calculation: Based on the timestamp sequence, calculate the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals; based on the standard time interval T_idea1 of N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals, calculate the time scaling factor K; where K=T_ideal / T_local.

[0028] In some embodiments, the actual cumulative time T_local of the local clock over N standard second pulse intervals is calculated using the following formula: T_local = C N -C0, where C0 and C N These are the first and last timestamps of the timestamp sequence, respectively.

[0029] In some implementations, the actual cumulative time T_local is obtained by weighted averaging of multiple consecutive timestamp intervals in the timestamp sequence. Wherein, the timestamp sequence (C0, C1, C2, ..., C...)... N The corresponding timestamp intervals are: C2-C1, C3-C2, ..., C N -C N-1 In practice, the timestamp intervals in the middle section can be given higher weights to further optimize the filtering effect.

[0030] The time scaling factor K is essentially the ratio of the standard time elapsed rate to the local clock time elapsed rate. If K > 1, it means the local clock is too slow and needs to be "speeded up"; if K < 1, it means the local clock is too fast and needs to be "speeded down".

[0031] S3. Rate scaling correction: The time scaling factor K is used as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time t_corrected.

[0032] In some embodiments, the corrected time t_corrected is calculated using the following formula: t_corrected=(C_current-C_base)×K+T_base; Where C_base is the count value of the onboard computer's local clock when the standard second pulse arrives at the calibration reference point; T_base is the standard time of the corresponding standard second pulse; C_current is the count value of the current local clock to be calibrated; and K is the time scaling factor. For example, C_base can be C1, and T_base corresponds to the absolute number of seconds of the first pulse. C_current can be the count value of any local clock to calibrate any local clock count value.

[0033] This step ensures that the rate of time elapsed, starting from time C_base, is uniformly scaled to the standard rate. The essence of this correction method lies in multiplicative scaling. Unlike the "direct time reset" method, which directly sets the local time to the standard time, this embodiment adjusts the elapsed rate of the entire time interval from the correction reference point C_base to C_current uniformly by multiplying by a time scaling factor K. This guarantees that the corrected timeline is continuous, smooth, and without abrupt changes, which is crucial for GNC algorithms that rely on time continuity.

[0034] In some embodiments, step S4 is also included: S4. Sliding Update: To adapt to changes in clock drift, this embodiment uses a sliding window method to update the time scaling factor. For example, each time a new second pulse is received, the earliest one is discarded, and the time scaling factor K is recalculated based on the updated timestamp sequence, thus achieving dynamic and smooth updates of the time scaling factor K.

[0035] In some embodiments, the calculated new time scaling factor K_new may not be used immediately, but instead is subjected to a first-order low-pass filter (e.g., K=α×K_old+(1-α)×K_new) with the time scaling factor K_old of the previous cycle, making the coefficient change smoother and further suppressing high-frequency jitter in the output time.

[0036] Compared to existing technologies based on single-pulse measurement, which are susceptible to signal jitter and instantaneous errors, exhibit poor short-term stability and struggle to provide uniform and accurate timestamps within the millisecond-level period required by the GNC algorithm, the on-board synchronization time synchronization method in this embodiment not only ensures time continuity but also overcomes the shortcomings of poor short-term stability. Without introducing additional hardware, it utilizes a simple algorithm to generate high-quality on-board time that is both seamless, accurate over the long term, and uniform in the short term. The following description, in conjunction with the accompanying drawings, illustrates the time synchronization logic of this embodiment's on-board synchronization time synchronization method, which involves "calculating the multiplicative scaling factor based on multi-pulse averaging and performing continuous rate scaling."

[0037] Figure 2 In the diagram, the onboard computer's local clock—the raw timing generated by its internal crystal oscillator without external standard time correction—is represented by a red line perpendicular to the time axis. When using the onboard computer's local clock for calculations, its time scale must be converted to a dynamic time scale. Introducing a time scaling factor K not only ensures that the onboard time acquired at any time is synchronized with the dynamic system but also compensates for crystal oscillator drift within the onboard computer itself.

[0038] The following describes the calculation process of the time scaling factor K in a specific embodiment, with reference to the accompanying drawings. The GNSS receiver in the spacecraft system generates a continuous, high-precision pulse-per-second (PPS) signal, represented by the black line perpendicular to the time axis in the figure. The PPS signal is used as the absolute measure for time synchronization. The onboard computer's local clock uses SelfTime for timing. For example, when the first PPS signal (PPS1) arrives, the count value is SelfTime_PPS0, corresponding to timestamp C0; when the eleventh PPS signal (PPS11) arrives, the count value is SelfTime_PPS10, corresponding to timestamp C0. 10 The time scaling factor K is calculated by measuring the local clock time length between consecutive second pulse intervals. The formula is: K = N seconds / (C N -C0), where C0 and C N These are the local clock count values ​​when the 1st and (N+1)th second pulses arrive, respectively. For example, for 10+1 second pulses (i.e., a 10-second pulse interval), K = 10 / (C...). 10 -C0).

[0039] During calibration, the formula is used: Onboard Time = GNSS integer seconds + (Local clock count of the current onboard computer to be calibrated - Local clock count of the onboard computer corresponding to the current PPS) × K. Here, the GNSS integer seconds is the standard time corresponding to the current PPS; this standard time is the time information sent by GNSS to the onboard computer for each PPS. The introduction of K linearly scales the local clock time scale to the standard time scale.

[0040] The average of 10 PPS values ​​can smooth out PPS fluctuations caused by dynamics. Under continuous and normal PPS output, the time scaling factor is updated once per second using a sliding window.

[0041] This application treats the external second pulse as a long-term, absolutely accurate time scale. By calculating its "average" mapping relationship on the local clock, a "time scaling factor" is obtained, thereby performing a global, uniform scaling correction on the local clock. The time scaling factor K is defined for multiplication operations. The physical meaning of K = T_idea1 / T_1ocal is the ratio of the "standard time elapsed rate" to the "local clock time elapsed rate." During correction, multiplying by this ratio "stretches" or "compresses" the current local clock to be corrected onto the standard time scale. Therefore, using the time scaling factor as a multiplication factor is the fundamental reason why this method achieves "continuous uniform correction" rather than "jumps."

[0042] The on-board time synchronization method in this embodiment is based on the core idea of ​​calculating the time scaling factor using N consecutive second pulses. It employs the time scaling factor to perform multiplication operations on the current clock count to obtain the corrected time. A dynamic calibration mechanism for updating the time scaling factor using a sliding window is designed, which simultaneously ensures the continuity and short-term uniformity of time, moderate and controllable computational complexity, and performance close to that of hardware solutions, while avoiding the high cost of hardware solutions.

[0043] Based on the same inventive concept, this embodiment also provides an on-board time synchronization system based on second pulses, including: The signal acquisition module acquires N+1 consecutive external standard second pulses and records the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1. The caching module is used to store timestamp sequences. If the timestamp sequence is not continuous, recording will start again. The coefficient calculation module calculates the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals based on the timestamp sequence; and calculates the time scaling factor K based on the standard time interval T_idea1 of N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals; where K=T_ideal / T_local. The rate scaling correction module uses the time scaling factor as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time. The sliding update module uses a sliding window approach to discard the oldest second pulse each time a new second pulse is received. The trigger coefficient calculation module recalculates the time scaling factor K based on the updated timestamp sequence.

[0044] For details on the logic control components of each module, please refer to the previous section on on-board time synchronization control methods; they will not be repeated here.

[0045] Obviously, the above embodiments are merely examples to clearly illustrate the technical solutions of the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art can make other variations or modifications based on the above description. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the claims of the present invention.

Claims

1. A satellite time synchronization method based on second pulses, characterized in that, The method includes the following steps: S1. Signal Acquisition and Recording: Acquire N+1 consecutive external standard second pulses and record the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1; S2. Coefficient Calculation: Based on the timestamp sequence, calculate the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals; based on the standard time interval T_idea1 of the N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals, calculate the time scaling factor K; where K = T_ideal / T_local; S3. Rate scaling correction: The time scaling factor is used as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time.

2. The on-board time synchronization method based on second pulses according to claim 1, characterized in that, In step S2, the actual cumulative time T_local of the local clock within N standard second pulse intervals is calculated using the following formula: T_local=C N -C0; Among them, C0, C N These are the first and last timestamps of the timestamp sequence, respectively.

3. The on-board time synchronization method based on second pulses according to claim 1, characterized in that, The actual cumulative time T_local of the local clock over N standard second pulse intervals is calculated using the following method: The actual cumulative time T_local is obtained by weighted averaging of multiple consecutive timestamp intervals in the timestamp sequence.

4. The on-board time synchronization method based on second pulses according to claim 1, characterized in that, In step S3, the timing rate of the onboard computer's local clock is continuously and uniformly scaled using the following formula to generate the corrected time t_corrected: t_corrected=(C_current-C_base)×K+T_base; Where C_base is the count value of the onboard computer's local clock when the standard second pulse arrives at the calibration reference point, T_base is the standard time of the corresponding standard second pulse, C_current is the count value of the current local clock to be calibrated, and K is the time scaling factor.

5. The on-board time synchronization method based on second pulses according to claim 1, characterized in that, The value of N is dynamically adjusted according to the spacecraft system's requirements for accuracy and response speed; when the clock is unstable, the value of N is greater than the value of N when it is stable.

6. The on-board time synchronization method based on second pulses according to claim 1, characterized in that, The value of N is 10.

7. The on-board time synchronization method based on second pulses according to any one of claims 1 to 6, characterized in that, The method further includes the following steps: The time scaling factor is updated using a sliding window method: each time a new second pulse is received, the earliest one is discarded, and the time scaling factor K is recalculated based on the updated timestamp sequence.

8. The on-board time synchronization method based on second pulses according to claim 7, characterized in that, The calculated new time scaling factor K_new is then used after being subjected to a first-order low-pass filter with the time scaling factor K_old from the previous cycle.

9. A satellite time synchronization system based on second pulses, characterized in that, The system includes: The signal acquisition module acquires N+1 consecutive external standard second pulses and records the count value corresponding to the local clock of the onboard computer when each second pulse arrives, forming a timestamp sequence; where N>1. The caching module is used to store timestamp sequences. If the timestamp sequence is not continuous, recording will start again. The coefficient calculation module calculates the actual cumulative time T_loca1 of the local clock within N standard second pulse intervals based on the timestamp sequence; and calculates the time scaling factor K based on the standard time interval T_idea1 of the N standard second pulses and the actual cumulative time T_local of the local clock within N standard second pulse intervals; where K = T_ideal / T_local. The rate scaling correction module uses the time scaling coefficient as a multiplication factor to continuously and uniformly scale the timing rate of the onboard computer's local clock to obtain the corrected time.

10. The on-board time synchronization system based on second pulses according to claim 9, characterized in that, The system also includes: The sliding update module uses a sliding window approach to discard the oldest second pulse each time a new second pulse is received. The trigger coefficient calculation module recalculates the time scaling factor K based on the updated timestamp sequence.

Images