A method for autonomous space-based maintenance of Beidou satellite-borne clock based on space station frequency reference

Abstract

The present invention discloses a method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference, relating to the field of time and frequency control technology. The method comprises: obtaining raw clock error data from a space station atomic clock and a BDS-3 satellite-borne clock and preprocessing the raw clock error data; performing sliding interpolation processing on the preprocessed clock error data sequence to obtain a continuous clock error data sequence; performing digital noise compensation on the continuous clock error data sequence; constructing a Kalman filter parameter estimation model using the compensated clock error data sequence to obtain the BDS-3 satellite-borne clock frequency steering amount for the next steering cycle; and performing frequency steering operations on the BDS-3 satellite-borne clock for the next steering cycle using the BDS-3 satellite-borne clock frequency steering amount. The present invention can reduce the dependence of the BDS-3 satellite-borne clock on the ground system, enabling it to maintain its autonomous operation in orbit for a longer period of time.

Description

Technical Field

[0001] The present invention belongs to the technical field of time and frequency control, and in particular relates to a method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference. Background Art

[0002] The BeiDou-3 (BDS-3) global satellite navigation system constellation consists of 24 medium earth orbit satellites, three geosynchronous satellites, and three inclined geosynchronous orbit satellites. The satellites carry a variety of onboard clocks, including rubidium and hydrogen atomic clocks. These clocks serve as onboard frequency references and are crucial satellite payloads. Their performance directly determines the quality of navigation, positioning, and timing services. Prior to the launch of the high-precision time and frequency experiment cabinet on the Chinese space station, the satellite-borne clocks, as the best frequency sources in space, relied primarily on high-precision remote comparison technology via satellite-to-ground and intersatellite links for performance evaluation and calibration.

[0003] Satellite-to-ground comparison allows real-time monitoring of the status of onboard clocks. When time or frequency accuracy falls short of expectations, frequency or phase modulation commands are issued to compensate. However, due to the limited distribution of overseas monitoring stations for the BDS-3 global satellite navigation system, most non-geosynchronous orbit satellites have intermittent ground visibility, making it difficult to rely on ground-based time and frequency synchronization for onboard clocks. To address this issue, BDS-3 satellites are equipped with a Ka-band intersatellite link comparison payload, which maintains comparison with ground stations via a cascade of satellite-to-ground and intersatellite links. However, cumulative errors in these multiple comparison links, and the potential for systematic errors between different links, can affect the synchronization accuracy of onboard clocks. With the launch of the China Space Station's High-Precision Time and Frequency Experiment Cabinet, the high-precision time and frequency system, equipped with an active hydrogen atomic clock, a cold atomic microwave clock, and a cold atomic strontium optical clock, has improved the daily stability of my country's space frequency reference to the order of E-17, far exceeding the long-term stability of the current BDS-3 onboard clocks. Combining the Ka-band two-way time and frequency comparison payloads of the space station and the BDS-3 satellite essentially allows the use of the space station to enhance the long-term stability of the BeiDou satellite-borne clock and improve the autonomous maintenance capability of the space-based time and frequency reference. However, there is currently no research or solution for operating the BDS-3 satellite-borne clock using a space optical clock.

[0004] One existing method uses a satellite-to-ground comparison link, employing high-precision remote comparison technology to compare with the ground. If the time or frequency accuracy falls short of expectations, frequency or phase modulation commands are injected to compensate, completing the BDS-3 onboard clock performance evaluation and calibration. Another method cascades the satellite-to-ground comparison link with an intersatellite comparison link, employing high-precision remote comparison technology to compare with the ground. If the time or frequency accuracy falls short of expectations, frequency or phase modulation commands are injected to compensate, completing the BDS-3 onboard clock performance evaluation and calibration.

[0005] However, the premise for monitoring the status of onboard clocks through the satellite-to-ground comparison link is that the BDS-3 satellites are visible to the ground stations. However, due to the small number of overseas monitoring stations of the BDS-3 global satellite navigation system, most non-geosynchronous orbit satellites have discontinuous ground visibility problems, making it difficult to ensure the continuity of the comparison between BDS-3 satellites and ground stations. The method of cascading the satellite-to-ground comparison link and the inter-satellite comparison link avoids the problem of discontinuous comparison data of the method based on the satellite-to-ground comparison link. However, multi-level comparison links will introduce new cumulative errors, and there may be systematic errors between different links, affecting the synchronization accuracy of the onboard clocks.

[0006] In summary, the current performance evaluation of satellite-borne clocks and time synchronization are highly dependent on the ground system and the accuracy of satellite-ground comparisons. Under conditions such as adversarial conditions, the ground system may find it difficult to ensure service continuity and reliability, which may affect the service performance of the Beidou system. Summary of the Invention

[0007] In order to solve the above problems existing in the prior art, the present invention provides a method for autonomous space-based maintenance of Beidou satellite-borne clocks based on a space station frequency reference. The technical problem to be solved by the present invention is achieved through the following technical solutions:

[0008] The present invention provides a method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference, comprising:

[0009] S1: Acquire the raw clock difference data between the space station atomic clock and the onboard clocks of each BDS-3 satellite through the inter-satellite link, and pre-process the raw clock difference data within a preset acquisition period to obtain a pre-processed clock difference data sequence;

[0010] S2: performing sliding interpolation processing on the pre-processed clock error data sequence to obtain a continuous clock error data sequence;

[0011] S3: performing digital noise compensation on the continuous clock error data sequence to obtain a compensated clock error data sequence;

[0012] S4: constructing a Kalman filter parameter estimation model using the compensated clock error data sequence, and obtaining the BDS-3 satellite-borne clock frequency steering value for the next steering cycle using the Kalman filter parameter estimation model;

[0013] S5: Using the BDS-3 onboard clock frequency control value, perform frequency control operations on the onboard clocks of each BDS-3 satellite in the next control period.

[0014] Compared with the prior art, the present invention has the following beneficial effects:

[0015] 1. Addressing the research gap in using space optical clocks to control satellite-borne clocks, the present invention proposes a method for autonomous space-based maintenance of Beidou satellite-borne clocks based on the space station frequency reference, providing a feasible approach for establishing and autonomously maintaining space-based time and frequency references. The application of the method of the present invention can effectively reduce the dependence of the BDS-3 satellite-borne clock on the ground system, enabling it to have a longer-term autonomous on-orbit maintenance capability. At the same time, the method can significantly improve the long-term stability of the BDS-3 satellite-borne clock. Specifically, the 1d stability of the BDS-3 satellite-borne clock can be improved by about half an order of magnitude, and the 10d stability can be improved by about two orders of magnitude.

[0016] 2. To address the problem of data fragmentation between the comparison between the space station atomic clock and the BDS-3 satellite onboard clock due to discontinuous visibility, the noise-adaptive digital compensation interpolation method proposed in this paper (steps S2 and S3) can obtain continuous clock error data that is more consistent with the actual operation of the space station atomic clock. A Kalman filter parameter estimation model is constructed based on the compensated clock error data sequence. Combined with the predicted clock error and frequency deviation values ​​for the next steering cycle, the frequency steering amount for each BDS-3 satellite onboard clock for the next steering cycle is calculated. This frequency steering is then performed on each BDS-3 satellite onboard clock, enabling it to maintain autonomous on-orbit for a longer period of time.

[0017] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference provided by an embodiment of the present invention;

[0019] Figure 2 This is another flow chart of a method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference provided by an embodiment of the present invention;

[0020] Figure 3 This is a diagram showing the stability results of a BDS-3 M22 satellite onboard clock before and after steering, provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0021] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following is a detailed description of a Beidou satellite-borne clock space-based autonomous maintenance method based on a space station frequency reference proposed in accordance with the present invention, in combination with the accompanying drawings and specific implementation methods.

[0022] The aforementioned and other technical contents, features, and effects of the present invention are clearly presented in the following detailed description of the specific embodiments in conjunction with the accompanying drawings. Through the description of the specific embodiments, a deeper and more specific understanding of the technical means and effects adopted by the present invention to achieve the intended purpose can be obtained. However, the accompanying drawings are provided for reference and illustration purposes only and are not intended to limit the technical solutions of the present invention.

[0023] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations are intended to cover non-exclusive inclusion, such that an article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the article or device comprising the element.

[0024] In order to solve the problem that the performance evaluation and calibration of satellite-borne clocks are currently highly dependent on ground systems and the accuracy of satellite-to-ground comparisons, and at the same time fill the gap in the use of space optical clocks to control satellite-borne clocks, the present invention provides a method for the space-based autonomous maintenance of Beidou satellite-borne clocks based on the space station frequency reference. Taking the E-17-level high-precision time and frequency system carried by the space station as a reference, by applying technologies such as a noise adaptive digital compensation interpolation algorithm and a space station-based satellite-borne clock control method, a new solution is provided for the establishment and autonomous maintenance of space-based time and frequency references.

[0025] The present invention's method for autonomous space-based maintenance of Beidou satellite-borne clocks utilizes the intersatellite link payload between the space station and the BDS-3 satellite to obtain clock error data between the BDS-3 satellite-borne clocks and the space station's atomic clock. To address the problem of clock error data fragmentation caused by discontinuous visibility between the BDS-3 satellites and the space station, a noise-adaptive digital compensation interpolation algorithm is proposed to improve data continuity. By establishing a Kalman filter parameter estimation model, the clock error of the next steering cycle is predicted and the corresponding frequency steering amount is calculated, achieving space-based autonomous maintenance of the BDS-3 satellite-borne clocks. Specifically, see Figure 1 and Figure 2 The Beidou satellite-borne clock autonomous space-based maintenance method of this embodiment specifically includes the following steps:

[0026] S1: The original clock difference data between the space station atomic clock and the BDS-3 satellite onboard clock are obtained through the intersatellite link, and the original clock difference data within the preset acquisition period are preprocessed to obtain the preprocessed clock difference data sequence.

[0027] Both the space station and the BDS-3 satellite are equipped with Ka-band intersatellite link payloads. These payloads utilize phased array antennas to transmit and receive signals. Signals propagate at the speed of light in space. By measuring the round-trip time between the space station and the BDS-3 satellite, the distance between them can be calculated. Since signal propagation time is related to the satellite clock error, measuring and processing this round-trip time allows the clock error between the satellites to be isolated.

[0028] It should be noted that most non-geosynchronous orbit satellites have the problem of discontinuous ground visibility. Therefore, the space station and the BDS-3 satellite can only exchange data when they are visible, but not when they are not visible. In other words, the clock difference data between the space station atomic clock and the onboard clocks of each BDS-3 satellite cannot be obtained when they are not visible. Therefore, the collected raw clock difference data is discontinuous and is stored according to a preset collection period (usually in days, i.e. 24 hours).

[0029] In this example, an intersatellite comparison link is used to acquire clock difference data between the space station's atomic clock and the BDS-3 satellite's onboard clock. The sampling interval is 30 seconds, meaning that a clock difference data point is collected every 30 seconds. As mentioned above, the space station and BDS-3 satellite can only exchange data when they are visually connected. Therefore, the collected clock difference data is discontinuous, forming a discontinuous clock difference data sequence.

[0030] After collecting raw clock error data for a preset time period, this data needs to be subjected to gross error detection and elimination. Specifically, the MAD method is used to filter and eliminate outliers such as occasional gross errors from the collected raw clock error data. In this embodiment, the preset time period, i.e., the preset collection period, is 24 hours. After collecting 24 hours of raw clock error data, the clock error data sequence collected within that 24 hours for the space station and the current BDS-3 satellite is considered as a whole and subjected to outlier screening and elimination.

[0031] The specific definition of MAD method is:

[0032] , (1)

[0033] in, represents the median absolute deviation, is the function for taking the median, is a sequence of original clock error data, is the first The measured value of the clock difference data. Calculate That is to find the median of the data's absolute deviation from the median, and then use formula (2) to screen outliers.

[0034] , (2)

[0035] That is, if the data deviates from the median More than 3 times , the current measurement value is considered to be an outlier, and the above judgment is performed on all clock difference data in the original clock difference data sequence, thereby eliminating all outliers and obtaining the preprocessed clock difference data sequence.

[0036] S2: Perform sliding interpolation processing on the preprocessed clock error data sequence to obtain a continuous clock error data sequence.

[0037] Specifically, a quadratic fit is performed on the clock error data sequence preprocessed in step S1 to construct a quadratic polynomial model of the space station atomic clock error. Subsequently, a sliding interpolation process is performed on the discontinuous segments in the preprocessed clock error data sequence based on the quadratic polynomial model to obtain a continuous clock error data sequence. The expression of the quadratic polynomial model is:

[0038] (3)

[0039] in, The square fitting of the space station atomic clock and the BDS-3 satellite onboard clock is shown in Figure 2. Clock difference data at the time, It represents the initial time difference between the space station atomic clock and the BDS-3 satellite onboard clock. Indicates the frequency deviation of the BDS-3 satellite onboard clock relative to the space station atomic clock. Indicates the frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock. for The noise of the BDS-3 satellite onboard clock relative to the space station atomic clock at this moment.

[0040] Subsequently, the quadratic polynomial model is used to perform sliding interpolation processing on the discontinuous segments in the preprocessed clock error data sequence to obtain a continuous clock error data sequence.

[0041] Specifically, set the window length of the sliding window w and step length s , take all valid data in the window; use the quadratic polynomial model to perform quadratic fitting on the data in the window and solve the parameters , and then substitute Calculation time The corresponding interpolation results are obtained, and the above process is repeated until all discontinuous segments of the preprocessed clock error data sequence are covered.

[0042] It should be noted that the sliding interpolation processing in this step is only for the discontinuous segments of the pre-processed clock error data sequence, and no sliding interpolation processing is performed on the received real clock error data.

[0043] S3: Perform digital noise compensation on the continuous clock error data sequence after the sliding interpolation processing to obtain a compensated clock error data sequence.

[0044] Since the noise of different types of atomic clocks is shown in the Sigma-Tau diagram (σ- The Sigma-Tau graph (σ- Figure ) is a tool for analyzing the frequency stability of atomic clocks by plotting the Allan variance (Allan Variance) versus integration time The logarithmic relationship between and can intuitively demonstrate the noise characteristics of the atomic clock.

[0045] Specifically, step S3 of this embodiment includes:

[0046] S3.1: Select a data segment with a first segment length of not less than 600 seconds from the clock error data sequence preprocessed in step S1 as a clock error data sample, and calculate the Allan variance of the clock error data sample.

[0047] Allan variance is a statistical measure of the instability of the atomic clock frequency, which is reflected in the integration time The calculation formula of Allan variance is an existing formula and will not be repeated here.

[0048] S3.2: Based on the relationship between Allan variance and power spectrum noise, the corresponding noise index is obtained μ .

[0049] Specifically, the relationship between Allan variance and power spectrum noise can be expressed as:

[0050] , (4)

[0051] in, represents the integration time, μ represents the noise figure, A ( μ ) represents the noise level, Indicates integration time The corresponding Allan variance.

[0052] Substitute the Allan variance of the clock error data sample calculated in step S3.1 into the above formula (4) to obtain the noise index μ.

[0053] S3.3: Determine the noise type affecting the space station atomic clock based on the correspondence between the noise index and the noise type, and construct a digital noise compensation model based on the noise type.

[0054] Please refer to Table 1, which shows the correspondence between common noise and its noise index. From Table 1, we can get the noise index. μ The corresponding noise type.

[0055] Table 1 Correspondence between common noise and its noise index

[0056]

[0057] After obtaining a clear noise type, a digital noise compensation model adapted to it is constructed. Then, based on the dynamic compensation coefficient adjustment mechanism of the digital noise compensation model, the continuous clock difference data sequence after the sliding interpolation processing in step S2 is digitally compensated for noise. After this processing, the interference of noise on the clock difference data can be significantly suppressed, and continuous clock difference data that is closer to the actual operating state of the atomic clock can be obtained.

[0058] In this embodiment, taking the BDS-3 M22 satellite as an example, calculations and table lookup verification show that the clock error between the space station atomic clock and the BDS-3M22 satellite's onboard clock is mainly affected by FM white noise and FM flicker noise, of which the influence of FM white noise accounts for about 90%. Therefore, this embodiment mainly focuses on digital noise compensation for FM white noise.

[0059] Taking into account the limitations of the visible period between the space station and the BDS-3 M22 satellite and the noise characteristics of their atomic clocks, it can be concluded that digital compensation of the white frequency-modulated noise of the onboard clock can be achieved using limited inter-satellite link comparison data.

[0060] Specifically, to construct a digital noise compensation model, we first need to generate a set of random numbers that obey a normal distribution, and then calculate the variance of this set of random numbers. The calculation steps are:

[0061] (1) Derive the relationship between random number variance and Allan variance, where Allan variance is defined as:

[0062] , (5)

[0063] in, For the n The clock difference data of the sampling points, Indicates the n + N The clock difference data of the sampling points, Indicates the The clock difference data of the sampling points, , is the sampling interval of the sampling point.

[0064] set up is a random number that conforms to the normal distribution, with a mean of 0 and a variance of ; FM white noise random number obeys the normal distribution, that is, ; Assume that the sampling interval is unit time, and the clock difference data of the first sampling point is , then the FM white noise random number can be approximately expressed as:

[0065] , (6)

[0066] Right now:

[0067] , (7)

[0068] Substituting formula (6) into formula (4) yields:

[0069] , (8)

[0070] in, It means to find the mathematical expectation.

[0071] Transform formula (8) to obtain the relationship between random number variance and Allan variance:

[0072] , (9)

[0073] (2) According to the relationship between random number variance and Allan variance, the random number variance is calculated by calculating the Allan variance of the typical points of the white noise of the atomic clock frequency modulation of the space station.

[0074] Specifically, based on previous analysis of various types of atomic clocks, we know the locations of typical points of FM white noise for each type of atomic clock. For example, after analysis, we found that the influence of FM white noise of atomic clocks is obvious at 300s. Therefore, 300s is the typical point of FM white noise, and the Allan variance at 300 seconds can be calculated based on the clock error data. Assume The typical point of FM white noise on the Allan variance diagram is value, It should The Allan variance corresponding to the value is The variance of the random number corresponding to the value It can be expressed as:

[0075] , (10)

[0076] (3) Generate random numbers based on the calculated random number variance Processing is performed to obtain a set of frequency series that conform to the characteristics of atomic clock frequency modulation white noise :

[0077] , (11)

[0078] in, Represents the variance of random numbers The arithmetic square root of .

[0079] The simulated FM white noise phase sequence is:

[0080] , (12)

[0081] in, express The frequency modulated white noise corresponding to the clock difference data at the moment, express The frequency modulated white noise corresponding to the clock difference data at the moment.

[0082] S3.4: Based on the dynamic compensation coefficient adjustment mechanism of the digital noise compensation model, the noise of the continuous clock error data sequence after sliding interpolation processing is digitally compensated to obtain the compensated clock error data sequence.

[0083] Specifically, the frequency modulated white noise phase sequence is added to the continuous clock error data sequence after sliding interpolation processing to obtain the compensated clock error data sequence :

[0084] , (13)

[0085] The above method can realize adaptive digital compensation of atomic clock frequency modulation white noise.

[0086] It should be noted that in order to preserve the true characteristics of the atomic clock as much as possible, the noise value of the effective clock error data segment in the simulated FM white noise phase sequence is set to 0. In other words, for the actual clock error data, that is, the value of the FM white noise compensation model corresponding to the originally obtained discontinuous clock error data is set to 0, and the FM white noise of the clock error data obtained by sliding interpolation is expressed as the above formula (12).

[0087] S4: Use the compensated clock error data sequence to construct a Kalman filter parameter estimation model, and use the Kalman filter parameter estimation model to obtain the BDS-3 satellite clock frequency control value for the next control cycle.

[0088] The Kalman filter is an algorithm that uses linear system state equations to estimate system states by inputting and outputting observation data. It is based on the recursive optimal estimation theory and is suitable for handling prediction and filtering problems of dynamic systems with noise. This step constructs a Kalman filter parameter estimation model based on the clock error data sequence compensated in step S3. The Kalman filter parameter estimation model is used to predict the initial clock error (( a ), the frequency deviation of the BDS-3 satellite onboard clock relative to the space station atomic clock ( b ) and the frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock ( c ) to predict the clock error for the next steering cycle. The predicted clock error is then combined with the frequency deviation to calculate the BDS-3 onboard clock frequency control for the next steering cycle. The steering cycle is determined by the expected value of the accumulated phase difference per steering cycle.

[0089] Specifically, step S4 of this embodiment includes:

[0090] S4.1: Construct the state vector and measurement vector of the Kalman filter parameter estimation model and initialize the parameters of the state vector and measurement vector. The state vector includes the initial clock difference between the space station atomic clock and the BDS-3 satellite onboard clock, and the frequency deviation and frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock.

[0091] The three parameters to be estimated The state vector that forms the Kalman filter parameter estimation model ,Right now ,in, Represents the atomic clock The clock error, frequency deviation and frequency drift of the driving cycle are calculated. The Kalman filter parameter estimation model is obtained from the driving cycle to The recursive formula for a driving cycle is:

[0092] , (14)

[0093] in, Indicates the A driving cycle, is a 3×3 dimensional state transition matrix, is the three-dimensional noise input to the Kalman filter parameter estimation model. The noise obeys a normal distribution with a mean of zero and is uncorrelated in time. Indicates the The state vector of a driving cycle.

[0094] Measurement vector It can be represented by the following matrix:

[0095] (15)

[0096] in, Indicates the driving cycles, measuring vector In the Measurements through the measurement matrix With the state vector Linear association, For the White noise of a driving cycle.

[0097] In this embodiment, the state transfer matrix in the Kalman filter parameter estimation model is and measurement matrix They are:

[0098] ,

[0099] , (16)

[0100] in, represents the state transition matrix, It should be noted that the state transfer matrix and the measurement matrix remain unchanged during the calculation process of the Kalman filter parameter estimation model, that is, they remain unchanged in each driving cycle.

[0101] State noise variance matrix and the measurement noise variance matrix It can be represented by the phase white noise, frequency white noise, frequency random walk noise and measurement noise variance of the clock error respectively:

[0102]

[0103] , (17)

[0104] in, is the sampling interval, and They are the variances of the phase white noise, frequency white noise, frequency random walk noise and measurement noise of the clock error, which can be solved by the Hadamard variance equation:

[0105] , (18)

[0106] Furthermore, the state vector and the error variance matrix The initial value of is set by experience. In this embodiment, the state vector The initial value is , since the state vector contains three states, the error variance matrix The initial value of needs to be determined through at least three measurements, specifically:

[0107] , (19)

[0108] in

[0109] ,

[0110] ,

[0111] ,

[0112] ,

[0113] ,

[0114] ,

[0115] in, and Respectively represent the time interval from the first measurement to the second measurement and the time interval from the second measurement to the third measurement. In this embodiment and The sampling time interval is taken.

[0116] S4.2: Use the Kalman filter parameter estimation model to predict the state vector of the space station atomic clock for the next steering cycle.

[0117] Specifically, this step uses the Kalman filter parameter estimation model to predict the state vector of the space station atomic clock for the next driving cycle based on the state vector and measurement vector of the constructed Kalman filter parameter estimation model, as well as the parameters of the initialized state vector and measurement vector. The specific processing process of the Kalman filter parameter estimation model is the same as that of the existing Kalman filter, which will not be repeated here.

[0118] S4.3: Calculate the BDS-3 satellite-borne clock frequency control value for the next control cycle using the state vector of the space station atomic clock for the next control cycle.

[0119] In this step, the clock error prediction value for the next driving cycle is completed based on the estimated state parameter values, namely the initial time clock difference between the space station atomic clock and the BDS-3 satellite onboard clock, the frequency deviation and frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock. ,Right now:

[0120] , (20)

[0121] in, represents the state vector predicted by the Kalman filter parameter estimation model, To control the cycle, Indicates transpose.

[0122] According to experience, the frequency control of satellite clocks can be divided into two cases:

[0123] (1) When the clock error prediction value of the next driving cycle has the same sign as the frequency deviation in the estimated state vector (i.e., both are positive or both are negative), the control quantity should contain both the phase component and the frequency component of the onboard clock. In order to ensure the continuity of the clock error data sequence, the phase control quantity should be converted into the frequency control quantity. Combining the clock error prediction value and the state parameter prediction value of the next driving cycle, the frequency control quantity corresponding to the next driving cycle of the BDS-3 satellite onboard clock is calculated. for:

[0124] , (twenty one)

[0125] (2) When the clock error prediction value of the next cycle and the frequency deviation in the estimated state vector have different signs (i.e. one is positive and the other is negative), the control amount only needs to consider the frequency component. Combining the clock error prediction value and the state parameter prediction value of the next control cycle, the frequency control amount corresponding to the next control cycle of the BDS-3 satellite onboard clock is calculated. for:

[0126] , (twenty two)

[0127] in, and , respectively, are the attenuation factors for the phase and frequency components of the frequency control value for the next driving cycle (set based on empirical values ​​to prevent overshoot). This embodiment, through multiple experiments, found that using this method to control the BDS-3 satellite onboard clock, when the Kalman filter parameter estimation model converges to a stable state, the single frequency adjustment is approximately on the order of E-13. Given that the accumulated phase difference per driving cycle is expected to be no more than 500 ps after steering, and to avoid other errors introduced by frequent steering, the final steering cycle was determined to be 1 hour.

[0128] S5: Use the onboard clock frequency control value to perform frequency control operations on the onboard clocks of each BDS-3 satellite in the next control cycle.

[0129] Based on the frequency steering amount of the satellite clock calculated in step S4, the widely used "ping-pong" frequency steering strategy is adopted to perform frequency steering operations on the satellite clocks of each BDS-3 satellite, thereby improving the on-orbit autonomous maintenance capability of the BDS-3 satellite satellite clocks. In order to avoid the adverse effects of excessive single frequency steering on the stability and other performance indicators of the BDS-3 satellite satellite clocks, the present invention sets a reasonable steering threshold to ensure that the frequency steering process is smooth and controllable. The stability of the BDS-3 satellite satellite clocks before and after the steering is calculated using Allan variance. Figure 3 As shown, the ordinate represents the Allan deviation, from Figure 3 It can be seen that after the control, the 1-day stability of the BDS-3 satellite's onboard clock is improved by about half an order of magnitude, and the 10-day stability is improved by about two orders of magnitude, and the performance of the onboard clock is significantly improved.

[0130] The present invention proposes a method for space-based autonomous maintenance of the Beidou satellite-borne clock based on the space station frequency reference, which provides a feasible way for the establishment and autonomous maintenance of space-based time and frequency reference. The application of the method of the present invention can effectively reduce the dependence of the BDS-3 satellite-borne clock on the ground system, so that it has a longer-term on-orbit autonomous maintenance capability. At the same time, the method can significantly improve the long-term stability of the BDS-3 satellite-borne clock. Specifically, the 1d stability of the satellite-borne clock can be improved by about half an order of magnitude, and the 10d stability can be improved by about two orders of magnitude. In response to the problem of fragmentation of comparison data caused by discontinuous visibility between the space station atomic clock and the BDS-3 satellite-borne clock, the noise adaptive digital compensation interpolation method (steps S2 and S3) proposed in the present invention obtains continuous clock difference data that is closer to the actual operation of the atomic clock. A Kalman filter parameter estimation model is constructed based on the compensated clock error data sequence. Combined with the clock error prediction value and frequency deviation prediction value of the next driving cycle, the frequency control amount of each BDS-3 satellite's onboard clock in the next driving cycle can be calculated. The frequency of each BDS-3 satellite's onboard clock can be controlled to enable it to have the ability to maintain autonomy in orbit for a longer period of time.

[0131] Another embodiment of the present invention provides a storage medium storing a computer program for executing the steps of the method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference as described in the above embodiments. Another aspect of the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor invokes the computer program in the memory, the steps of the method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference as described in the above embodiments are executed. Specifically, the integrated module implemented in the form of a software functional module can be stored in a computer-readable storage medium. The software functional module is stored in a storage medium and includes several instructions for causing an electronic device (which can be a personal computer, server, or network device, etc.) or a processor to execute some of the steps of the method described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0132] The above is a further detailed description of the present invention in conjunction with specific preferred embodiments, and the specific implementation of the present invention should not be considered to be limited to these descriptions. For those skilled in the art of the present invention, without departing from the concept of the present invention, several simple deductions or substitutions can be made, which should be considered to fall within the scope of protection of the present invention.

Claims

1. A method for autonomous space-based maintenance of a Beidou satellite-borne clock based on a space station frequency reference, characterized in that: include: S1: Acquire the raw clock difference data between the space station atomic clock and the onboard clocks of each BDS-3 satellite through the inter-satellite link, and pre-process the raw clock difference data within a preset acquisition period to obtain a pre-processed clock difference data sequence; S2: performing sliding interpolation processing on the pre-processed clock error data sequence to obtain a continuous clock error data sequence; S3: performing digital noise compensation on the continuous clock error data sequence to obtain a compensated clock error data sequence; S4: constructing a Kalman filter parameter estimation model using the compensated clock error data sequence, and obtaining the BDS-3 satellite-borne clock frequency steering value for the next steering cycle using the Kalman filter parameter estimation model; S5: Using the BDS-3 satellite-borne clock frequency control value, perform frequency control operations on the BDS-3 satellite-borne clocks in the next control period. Said S1 comprises: S1.1: Using the inter-satellite comparison link, obtain the original clock difference data between the space station atomic clock and the onboard clock of each BDS-3 satellite to form a set of discontinuous clock difference data sequences; S1.2: After 24 hours of data collection, use the MAD method to remove outliers in the clock error data sequence to obtain the pre-processed clock error data sequence between the space station atomic clock and the onboard clock of each BDS-3 satellite. The S2 includes: S2.1: Construct a quadratic polynomial model: ， in, The square fitting of the space station atomic clock and the BDS-3 satellite onboard clock is shown in Figure 2. Clock difference data at the time, a It represents the initial time difference between the space station atomic clock and the BDS-3 satellite onboard clock. Indicates the frequency deviation of the BDS-3 satellite onboard clock relative to the space station atomic clock. Indicates the frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock. for The noise of the BDS-3 satellite onboard clock relative to the space station atomic clock at this moment; S2.2: Using the quadratic polynomial model, perform sliding interpolation processing on the discontinuous segments in the preprocessed clock error data sequence to obtain a continuous clock error data sequence; The S3 includes: S3.1: Selecting a first data segment of at least 600 seconds in length from the clock error data sequence preprocessed in step S1 as a clock error data sample, and calculating the Allan variance of the clock error data sample; S3.2: Obtain a corresponding noise index based on the relationship between the Allan variance and the power spectrum noise; S3.3: Determine a noise type affecting the space station atomic clock based on the correspondence between the noise index and the noise type, and construct a digital noise compensation model based on the noise type; S3.4: Based on the dynamic compensation coefficient adjustment mechanism of the digital noise compensation model, the continuous clock difference data sequence after sliding interpolation processing is subjected to digital noise compensation to obtain a compensated clock difference data sequence.

2. The method for autonomous space-based maintenance of a BeiDou satellite-borne clock based on a space station frequency reference according to claim 1, characterized in that: The BDS-3 satellite is a BDS-3 M22 satellite. A digital noise compensation model is constructed according to the noise type, including: Construct a set of random numbers that follow a normal distribution , so that the relationship between the random number variance and the Allan variance satisfies: ， in, represents the Allan variance, Represents a constructed random number The variance of represents the integration time, N = / , is the sampling interval; According to the relationship between random number variance and Allan variance, the random number variance is calculated using the Allan variance calculation results of typical points of the white noise of the atomic clock frequency modulation of the space station; The random number generated based on the calculated random number variance Processing is performed to obtain a set of frequency modulated white noise phase series; The frequency modulated white noise phase sequence is added to the continuous clock difference data sequence after sliding interpolation processing to obtain a compensated clock difference data sequence.

3. The method for autonomous space-based maintenance of Beidou satellite-borne clock based on space station frequency reference according to claim 2, characterized in that: The random number generated based on the calculated random number variance After processing, a set of frequency modulated white noise phase series is obtained, including: The random number generated based on the calculated random number variance After processing, a set of frequency series that conforms to the white noise characteristics of the space station atomic clock frequency modulation is obtained. : ， in, express The variance of the random numbers corresponding to the values, express The arithmetic square root of Typical points of FM white noise on the Allan variance diagram value; Get the frequency modulated white noise phase sequence: ， in, express The frequency modulated white noise corresponding to the clock difference data at the moment, express The frequency modulated white noise corresponding to the clock difference data at the moment.

4. The method for autonomous space-based maintenance of a BeiDou satellite-borne clock based on a space station frequency reference according to claim 1, characterized in that: The S4 includes: S4.1: Construct the state vector and measurement vector of the Kalman filter parameter estimation model and initialize the parameters of the state vector and the measurement vector. The state vector includes the initial time clock difference between the space station atomic clock and the BDS-3 satellite onboard clock, and the frequency deviation and frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock. S4.2: Predicting the state vector of the space station atomic clock for the next steering cycle using the Kalman filter parameter estimation model; S4.3: Calculate the BDS-3 satellite-borne clock frequency control value for the next control cycle using the state vector of the space station atomic clock for the next control cycle.

5. The method for autonomous space-based maintenance of Beidou satellite-borne clock based on space station frequency reference according to claim 4, characterized in that: The S4.1 includes: The three parameters to be estimated The state vector forming the Kalman filter parameter estimation model ,Right now ,in, Represents the space station atomic clock The initial time clock error, frequency deviation and frequency drift of the driving cycle are used to obtain the Kalman filter parameter estimation model from the first driving cycle to The recursive formula for a driving cycle is: ， in, Indicates the A driving cycle, is a 3×3 dimensional state transition matrix, is the three-dimensional noise input to the Kalman filter parameter estimation model, Indicates the The state vector of a driving cycle; The measurement vector Expressed as: ， in, Indicates the A driving cycle, is the measurement matrix, For the White noise of a driving cycle; Set the state transfer matrix in the Kalman filter parameter estimation model and measurement matrix They are: ， ， in, represents the state transition matrix, represents the measurement matrix; Set the state noise variance matrix and the measurement noise variance matrix They are: ， ， in, is the sampling interval, and They are the phase white noise, frequency white noise, frequency random walk noise and measurement noise variance of the clock error.

6. The method for autonomous space-based maintenance of a BeiDou satellite-borne clock based on a space station frequency reference according to claim 4, characterized in that: The S4.3 includes: S4.31: Based on the estimated state parameter values, namely, the initial time difference between the space station atomic clock and the BDS-3 satellite onboard clock, the frequency deviation and frequency drift of the BDS-3 satellite onboard clock relative to the space station atomic clock, the clock error prediction value for the next driving cycle is obtained. : ， in, represents the state vector predicted by the Kalman filter parameter estimation model, To control the cycle, represents transpose; S4.32: When the clock error prediction value of the next driving cycle and the frequency deviation in the estimated state vector are both positive or negative, calculate the frequency control value corresponding to the next driving cycle of the BDS-3 satellite onboard clock for: ， When one of the clock error prediction value for the next driving cycle and the frequency deviation in the estimated state vector is positive and the other is negative, calculate the frequency control value corresponding to the next driving cycle of the BDS-3 satellite onboard clock. for: ， in, and They are the attenuation factors of the phase component and frequency component in the frequency control amount of the next control cycle respectively.

Images