A satellite and Loran timing data fusion method and related device

Abstract

The present invention belongs to the field of time synchronization and data processing, and discloses a method and related device for fusing satellite and Loran timing data. The method first truncates and removes the mean of the raw data to eliminate baseline offsets. Frequency domain transformation combined with singular value decomposition is used to reconstruct the frequency domain matrix. A dynamic threshold strategy is used to accurately identify and filter out periodic interference signals in the frequency domain. Time domain transformation is then used to retain the effective components. Finally, an adaptive Kalman filter algorithm is used to dynamically adjust the noise covariance matrix and observation model parameters to achieve dynamic weighted fusion of dual-source data. This method effectively overcomes the shortcomings of the traditional weighted averaging method, such as the insensitivity of fixed weights, the rigidity of Kalman filter parameters, and the insufficient global interference suppression of wavelet transforms. It significantly improves the ability to suppress non-stationary noise and periodic interference, enabling the fused data to combine the high precision of the satellite system with the anti-interference characteristics of the Loran system, ultimately achieving highly reliable and stable time synchronization performance.

Description

Technical Field

[0001] The present invention belongs to the technical field of time synchronization and data processing, and in particular relates to a satellite and Loran timing data fusion method and related devices. Background Art

[0002] In the field of time synchronization, accurate time acquisition is crucial for numerous industries and applications. Satellite timing systems, with their wide coverage and high precision, have become a key means of time acquisition, providing time services to numerous users worldwide. The Roland timing system, on the other hand, has attracted considerable attention for its strong anti-interference capabilities, maintaining stable operation even in certain complex environments. Both timing systems play important roles in the field of time synchronization, jointly supporting the need for time acquisition in diverse scenarios.

[0003] However, using either satellite or Loran timing systems alone presents significant limitations. While satellite timing systems offer significant advantages, they are susceptible to factors such as signal obstruction, multipath effects, and interference from the space environment, resulting in reduced time synchronization accuracy. While the Loran timing system offers strong anti-interference capabilities, it suffers from relatively low accuracy and limited signal coverage. To overcome the limitations of a single system, the idea of ​​fusing satellite and Loran timing data has emerged, hoping to leverage the strengths of both and compensate for their respective shortcomings. However, current data fusion methods struggle to effectively remove noise and significant periodic interference from the data in practice. For example, the weighted average method, as a basic fusion method, suffers from poor fusion accuracy and poor noise and periodic interference suppression due to its fixed weights, which cannot be adjusted in real time based on the data's authenticity. Fixed-parameter Kalman filtering, however, struggles to accurately describe data uncertainty in complex interference environments, as key parameters cannot be dynamically adjusted based on data characteristics, resulting in suboptimal filtering performance. Finally, the wavelet transform method, when fusion timing data, is inadequate for addressing periodic interference with global characteristics.

[0004] It can be seen that the existing data fusion method is difficult to ensure the accuracy and reliability of satellite and Loran timing fusion data, resulting in the inability to meet the needs of high-precision time synchronization. Summary of the Invention

[0005] The present invention provides a method and related device for fusing satellite and Loran timing data. This method performs frequency domain singular value decomposition fusion denoising on satellite and Loran timing data, removes obvious periodic terms, and uses adaptive Kalman filtering for fusion, effectively improving the accuracy and reliability of timing data to meet the needs of high-precision time synchronization.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for fusing satellite and Loran timing data, comprising:

[0008] The obtained satellite timing data sequence is sequentially subjected to data truncation and mean removal to obtain a first mean-removed data sequence; the obtained Loran timing data sequence is sequentially subjected to data truncation and mean removal to obtain a second mean-removed data sequence;

[0009] Performing frequency domain conversion, singular value decomposition, and reconstruction on the first de-meaned data sequence and the second de-meaned data sequence in sequence to obtain a reconstructed frequency domain matrix;

[0010] Based on a set threshold in the frequency domain, the reconstructed frequency domain matrix is ​​subjected to periodic interference signal removal and time domain conversion to obtain a first data sequence and a second data sequence; the first data sequence is a satellite timing data sequence after denoising and removing periodic terms, and the second data sequence is a Loran timing data sequence after denoising and removing periodic terms; the set threshold is calculated based on the reconstructed frequency domain matrix;

[0011] An adaptive Kalman filter algorithm is used to perform adaptive Kalman filter fusion on the first data sequence and the second data sequence to obtain a fused timing data sequence.

[0012] Furthermore, the step of sequentially performing data truncation and mean removal on the acquired satellite timing data sequence to obtain a first mean-removed data sequence; and sequentially performing data truncation and mean removal on the acquired Loran timing data sequence to obtain a second mean-removed data sequence comprises:

[0013] respectively intercepting valid data segments of the satellite timing data sequence and the Loran timing data sequence to obtain a first original data sequence and a second original data sequence, and making the first original data sequence and the second original data sequence have the same length;

[0014] The first original data sequence and the second original data sequence are subjected to mean removal processing respectively to obtain a first mean-removed data sequence and a second mean-removed data sequence. The specific formula is as follows:

[0015]

[0016] Where, represents the first mean-devalued data sequence; represents the second mean-devalued data series; represents a first original data sequence; represents the second original data sequence; represents the mean of the first original data series; represents the mean of the second original data series.

[0017] Furthermore, the first de-meaned data sequence and the second de-meaned data sequence are sequentially subjected to frequency domain conversion, singular value decomposition, and reconstruction to obtain a reconstructed frequency domain matrix, including:

[0018] Perform column vector conversion on the first de-meaned data sequence and the second de-meaned data sequence respectively to obtain a first column vector and a second column vector;

[0019] Performing frequency domain conversion on the first column vector and the second column vector respectively using fast Fourier transform to obtain a first frequency domain signal and a second frequency domain signal;

[0020] Combining the first frequency domain signal and the second frequency domain signal to obtain a frequency domain matrix;

[0021] Performing singular value decomposition on the frequency domain matrix to obtain a singular value matrix; the singular value matrix includes a left singular vector matrix, a singular value diagonal matrix and a right singular vector matrix;

[0022] Perform singular value screening on the singular value diagonal matrix to construct a new singular value diagonal matrix;

[0023] The frequency domain matrix is ​​reconstructed using the left singular vector matrix, the new singular value diagonal matrix and the right singular vector matrix to obtain a reconstructed frequency domain matrix.

[0024] Furthermore, before removing the periodic interference signal and performing time domain conversion on the reconstructed frequency domain matrix based on the set threshold in the frequency domain to obtain the first data sequence and the second data sequence, the method includes:

[0025] According to the maximum absolute value of the elements in the reconstructed frequency domain matrix, the set threshold is calculated and the formula is:

[0026]

[0027] Where, Indicates setting the threshold. Represents the reconstructed frequency domain matrix; Represents the maximum absolute value of the elements in the reconstructed frequency domain matrix.

[0028] Furthermore, based on the set threshold in the frequency domain, the reconstructed frequency domain matrix is ​​subjected to removal of periodic interference signals and time domain conversion to obtain the first data sequence and the second data sequence, including:

[0029] The elements in the reconstructed frequency domain matrix whose absolute values ​​are greater than the set threshold are set to 0 to remove the periodic interference signal from the reconstructed frequency domain matrix;

[0030] By using inverse fast Fourier transform, the first row and the second row of the reconstructed frequency domain matrix with the periodic interference signal removed are respectively converted into the time domain to obtain a first data sequence and a second data sequence.

[0031] Furthermore, the step of performing adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filter algorithm to obtain a fused timing data sequence includes:

[0032] Step 1: Obtain first data and second data at the current moment from the first data sequence and the second data sequence, and combine the first data and the second data into an observation vector; the first data is the satellite timing data after denoising and removing the periodic term, and the second data is the Loran timing data after denoising and removing the periodic term;

[0033] Step 2: Calculate the Kalman gain based on the pre-calculated prior covariance and the preset observation matrix; update the state estimate based on the pre-calculated prior state estimate, the Kalman gain, and the observation vector; output the updated state estimate as the fused timing data;

[0034] Step 3: Adaptively adjust the observation noise covariance based on the preset forgetting factor and the pre-calculated observation residual;

[0035] Steps 1, 2, and 3 are executed cyclically until the data corresponding to all moments in the first data sequence and the second data sequence are traversed, and the fused timing data corresponding to each moment is obtained to obtain a fused timing data sequence.

[0036] Further,

[0037] The step 1 includes:

[0038] Initialize the parameters of the adaptive Kalman filter; the parameters include the state transfer matrix, the observation matrix, the process noise covariance, the observation noise covariance, the forgetting factor, the state estimate and the covariance;

[0039] Before the step 2, the method includes:

[0040] The prior state estimate at the current moment is calculated based on the state transfer matrix and the state estimate at the previous moment; the prior covariance at the current moment is calculated based on the state transfer matrix, the covariance at the previous moment and the process noise covariance.

[0041] A satellite and Loran timing data fusion system, comprising:

[0042] A preprocessing module is used to sequentially perform data truncation and mean removal on the acquired satellite timing data sequence to obtain a first mean-removed data sequence; and sequentially perform data truncation and mean removal on the acquired Loran timing data sequence to obtain a second mean-removed data sequence;

[0043] a decomposition and reconstruction module, configured to sequentially perform frequency domain conversion, singular value decomposition, and reconstruction on the first de-meaned data sequence and the second de-meaned data sequence to obtain a reconstructed frequency domain matrix;

[0044] a periodic term removal module configured to remove periodic term interference signals and perform time domain conversion on the reconstructed frequency domain matrix based on a set threshold in the frequency domain to obtain a first data sequence and a second data sequence; the first data sequence is a satellite timing data sequence denoised and periodic term removed, and the second data sequence is a Loran timing data sequence denoised and periodic term removed; the set threshold is calculated based on the reconstructed frequency domain matrix;

[0045] The data fusion module is used to perform adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filtering algorithm to obtain a fused timing data sequence.

[0046] A satellite and Loran timing data fusion device, comprising:

[0047] memory for storing computer programs;

[0048] A processor is used to implement the steps of the above-mentioned satellite and Loran timing data fusion method when executing the computer program.

[0049] A computer-readable storage medium stores a computer program, wherein the computer program is used to implement the steps of the above-mentioned satellite and Loran timing data fusion method when executed by a processor.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] The present invention provides a method for fusing satellite and Loran timing data. The method first truncates and averages the raw data to eliminate baseline offsets. Frequency domain transformation combined with singular value decomposition is then used to reconstruct the frequency domain matrix. A dynamic threshold strategy is used to accurately identify and filter out periodic interference signals in the frequency domain. Time domain transformation is then used to retain the effective components. Finally, an adaptive Kalman filter algorithm is used to dynamically adjust the noise covariance matrix and observation model parameters to achieve dynamic weighted fusion of dual-source data. Frequency domain reconstruction extracts principal components through matrix decomposition, which, combined with threshold filtering, can separate noise and periodic interference. The adaptive Kalman filter dynamically optimizes parameters to match the noise statistics in complex interference environments by evaluating data residual characteristics in real time. This method effectively overcomes the shortcomings of the traditional weighted averaging method, such as fixed weight insensitivity, Kalman filter parameter rigidity, and insufficient global interference suppression by the wavelet transform. It significantly improves the ability to suppress non-stationary noise and periodic interference, enabling the fused data to combine the high precision of the satellite system with the anti-interference characteristics of the Loran system, ultimately achieving highly reliable and stable time synchronization performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 Waveform diagrams of raw satellite timing data and Loran timing data provided by an embodiment of the present invention;

[0053] Figure 2 Comparison waveforms of satellite timing data before and after noise and periodic terms are removed, provided by an embodiment of the present invention;

[0054] Figure 3 Comparison waveforms before and after removing noise and periodic terms from the Loran timing data provided by an embodiment of the present invention;

[0055] Figure 4 A comparison chart of satellite and Loran timing data fusion results provided by an embodiment of the present invention;

[0056] Figure 5 A flowchart of a method for fusing satellite and Loran timing data provided by an embodiment of the present invention;

[0057] Figure 6 A schematic structural diagram of a satellite and Loran timing data fusion system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to further understand the content of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0059] In order to better understand the present technical solution, the technical terms involved in this invention are explained as follows:

[0060] The adaptive Kalman filter algorithm is an improvement to the traditional Kalman filter algorithm. It aims to address the performance degradation of traditional Kalman filters when system model parameters are uncertain or when the statistical characteristics of noise are unknown or changing. It improves the accuracy and robustness of the filter by estimating and adjusting the statistical characteristics of system noise and measurement noise in real time.

[0061] Singular Value Decomposition (SVD): An important matrix decomposition method applicable to real or complex matrices of arbitrary shapes. It decomposes a matrix into the product of three matrices, revealing the inherent structural properties of the matrix.

[0062] Fast Fourier Transform (FFT): An algorithm for efficiently computing the discrete Fourier transform and its inverse transform. The Fourier transform is widely used in signal processing, image processing, data analysis, and other fields. It is used to convert signals from the time domain to the frequency domain and analyze the signal's frequency components.

[0063] Inverse Fast Fourier Transform (IFFT): This is the inverse of the Fast Fourier Transform (FFT), used to convert a frequency-domain signal back to the time-domain. The FFT converts a time-domain signal into a frequency-domain representation, while the IFFT restores the frequency-domain representation back to the original time-domain signal.

[0064] To effectively remove noise and significant periodic interference from the data and thereby improve the accuracy of the fused data, this embodiment provides a method for fusing satellite and Loran timing data. This method is based on Fourier transform, frequency domain matrix construction, singular value decomposition (SVD), singular value screening, and frequency domain matrix reconstruction to effectively separate the signal from the noise and achieve denoising. The method also uses adaptive Kalman filtering to achieve the final fusion of satellite and Loran timing data. The forgetting factor is used to dynamically adjust the observation noise covariance, adapting in real time to the differences in the characteristics of satellite and Loran timing data in different environments.

[0065] For example, Figure 5 As shown, this embodiment provides a method for fusing satellite and Loran timing data, including:

[0066] The obtained satellite timing data sequence is sequentially subjected to data truncation and mean removal to obtain a first mean-removed data sequence; the obtained Loran timing data sequence is sequentially subjected to data truncation and mean removal to obtain a second mean-removed data sequence;

[0067] Performing frequency domain conversion, singular value decomposition, and reconstruction on the first de-meaned data sequence and the second de-meaned data sequence in sequence to obtain a reconstructed frequency domain matrix;

[0068] Based on a set threshold in the frequency domain, the reconstructed frequency domain matrix is ​​subjected to periodic interference signal removal and time domain conversion to obtain a first data sequence and a second data sequence, where the first data sequence is a satellite timing data sequence denoised and periodic item removed, and the second data sequence is a Loran timing data sequence denoised and periodic item removed; the set threshold is calculated based on the reconstructed frequency domain matrix;

[0069] An adaptive Kalman filter algorithm is used to perform adaptive Kalman filter fusion on the first data sequence and the second data sequence to obtain a fused timing data sequence.

[0070] The data fusion method provided by this embodiment is described in more detail below with reference to the accompanying drawings:

[0071] This embodiment provides a method for fusing satellite and Loran timing data, and the specific steps are as follows:

[0072] S1. Data acquisition and preprocessing.

[0073] S11. Obtain satellite timing data sequence and Loran timing data sequence (i.e., satellite original data and Loran original data), and intercept valid data segments of the same length, which are respectively recorded as the first original data sequence. and the second original data sequence .

[0074] S12, yes and Perform mean removal processing to obtain the first mean-removed data sequence and the second mean-removed data sequence. The specific formula is:

[0075] (1)

[0076] Where, represents the first mean-devalued data sequence; represents the second mean-devalued data series; represents a first original data sequence; represents the second original data sequence; represents the mean of the first original data series; represents the mean of the second original data series.

[0077] This step can eliminate the DC component in the data, making it easier to process later. Figure 1 shown.

[0078] S2, frequency domain SVD fusion denoising.

[0079] S21, data conversion: convert the first mean-removed data sequence and the second mean-removed data sequence into column vector form, that is:

[0080]

[0081] (2)

[0082] Where, is the column vector form of the first mean-removed data sequence, that is, the first column vector, express;

[0083] is the column vector form of the second mean-removed data sequence, that is, the second column vector, express.

[0084] S22, frequency domain conversion: respectively and Perform fast Fourier transform (FFT) to obtain the first frequency domain signal and the second frequency domain signal .

[0085] The FFT transform formula of the N points involved is:

[0086] (3)

[0087] Where, is the nth sampling point of the discrete time signal, n is the time series index, n=0, 1, ..., N-1; represents the signal amplitude at the discrete time point n in the time domain;

[0088] is the kth sampling point of the frequency domain discrete signal; it means the frequency is Signal amplitude and phase information at (complex value);

[0089] N represents the number of sampling points of the discrete signal (i.e. the number of points in the FFT). The number of sampling points must be an integer power of 2 (according to the requirements of the fast algorithm) and is used to determine the frequency domain resolution: the frequency interval is ( is the time domain sampling frequency);

[0090] Indicates the frequency sequence index, k=0, 1, ..., N-1; corresponding to different frequency components, when k=0 corresponds to the DC component (frequency is 0), when k=1, 2, ..., N / 2-1 corresponds to the positive frequency component, when k=N / 2, ..., N-1 corresponds to the negative frequency component;

[0091] Represents the time domain index, corresponding to a discrete time point. Therefore, n and k traverse all sampling points in the time domain and frequency domain respectively to establish a mapping relationship between the time domain and the frequency domain.

[0092] Represents the rotation factor (also called butterfly factor), which is a complex sinusoidal basis function used for orthogonal projection of time domain signal and frequency domain basis function. Its physical meaning is: it represents the frequency Combination of sine and cosine waves (Euler's formula: ), its function is to extract the amplitude and phase of each frequency component in the signal by accumulating the product of all time domain sampling points and different frequency basis functions;

[0093] represents the imaginary unit, , used to represent the phase information of the sine basis function; θ represents a real number variable.

[0094] S23. Construct frequency domain matrix: combine two frequency domain signals into a frequency domain matrix :

[0095] (4)

[0096] S24, Singular Value Decomposition (SVD): Perform singular value decomposition on the frequency domain matrix M to obtain three matrices U, S, and V, that is, ,in, for The orthogonal matrix is ​​a left singular vector matrix, whose column vectors represent the main change direction of the data in the frequency domain; yes The diagonal matrix of is a singular value diagonal matrix, the diagonal elements (Singular value) Arranged from large to small, the size of the singular value reflects the energy or importance of the data in the direction of the corresponding singular vector. The larger the value, the more significant the data feature in that direction. yes An orthogonal matrix whose column vector is called the right singular vector matrix.

[0097] Solution method: calculation and :

[0098] calculate , this is a By solving The eigenvalues ​​and eigenvectors of can be used to obtain the right singular vector matrix , The eigenvalue of and singular values The relationship between them is: .

[0099] calculate , this is a A symmetric matrix. Solve The eigenvalues ​​and eigenvectors of can be obtained from the left singular vector matrix .

[0100] Determine the singular value matrix :according to or Calculate the singular value of the eigenvalue and arrange the singular values ​​in descending order to form a singular value matrix .

[0101] S25. Singular value screening: set the number of singular values ​​to be retained , construct a new singular value matrix ,Will Center front OK The elements of the column are set to The corresponding front OK Column elements, the rest of the elements are set to 0, that is, .

[0102] S26, reconstruct the frequency domain matrix: reconstruct the frequency domain matrix through the filtered singular value matrix to obtain the reconstructed frequency domain matrix In this process, the components corresponding to larger singular values ​​are retained, and the components corresponding to smaller singular values ​​are removed. Smaller singular values ​​often correspond to noise components in the data, thereby achieving denoising of the original frequency domain data.

[0103] S3. Remove obvious periodic terms.

[0104] S31, threshold calculation: calculate and reconstruct the frequency domain matrix The maximum absolute value of the element is multiplied by 0.01 to get the set threshold , the formula is:

[0105] (5)

[0106] S32, periodic term removal: reconstruct the frequency domain matrix The absolute value is greater than the set threshold The elements of are set to 0, that is, , thereby removing the obvious periodic interference, thereby removing the obvious periodic interference in the data.

[0107] S33, time domain conversion: Use the inverse fast Fourier transform (IFFT) to remove the periodic interference signal to reconstruct the frequency domain matrix The first and second rows of the data are converted back to the time domain to obtain the denoised and periodic terms removed signals of the satellite and Loran timing data, i.e., the first data sequence and the second data sequence The conversion formula is:

[0108] (6)

[0109] Among them, the inverse fast Fourier transform (IFFT) transformation formula of the N points involved is:

[0110] (7)

[0111] Represents the nth sampling point of the time domain discrete signal, that is, the denoised satellite timing data or Loran timing data , represents the recovered time domain signal, which is used for subsequent adaptive Kalman filter fusion;

[0112] Represents the kth sampling point of the frequency domain signal, corresponding to the reconstructed frequency domain matrix The first or second row of data (corresponding to satellite and Loran timing data, respectively) is the frequency domain signal after frequency domain SVD denoising and periodic term removal, and the frequency components corresponding to low energy noise and obvious periodic terms have been filtered out;

[0113] N The same as the number of FFT points, for reconstructing the frequency domain matrix The number of columns (i.e., the number of sampling points) must satisfy N is an integer power of 2 (to adapt to fast algorithms), which determines the time resolution of the time domain signal. N The larger it is, the denser the time domain sampling points are;

[0114] k Represents the frequency domain index, corresponding to the frequency component of the frequency domain signal (consistent with the definition in FFT).

[0115] In this embodiment, k Traverse and reconstruct the frequency domain matrix All frequency points of , including the retained effective frequency components (corresponding to large singular values) and the remaining components after removing noise;

[0116] n Represents the time domain index, corresponding to the restored time domain signal sampling point ( n =0, 1, ..., N -1). By n and k Mapping, restoring the signal after frequency domain processing to a time domain signal that can be used for time synchronization;

[0117] Indicates the inverse rotation factor, which is the same as the rotation factor in FFT Conjugate, physical meaning: by accumulating the product of all frequency domain sampling points and different frequency basis functions (including positive frequency phase information), the frequency domain signal is reconstructed into the time domain signal. j is an imaginary unit, ensuring that the phase information is correctly restored in the inverse transform;

[0118] It represents a normalization factor to ensure energy conservation between IFFT and FFT transformations. In this embodiment, this factor is introduced to make the amplitude of the reconstructed time domain signal consistent with the original signal, thereby avoiding energy scaling errors.

[0119] In this embodiment, the satellite and Loran timing data are de-noised and de-periodicized using the SVD algorithm. Figure 2 and Figure 3 shown.

[0120] S4, adaptive Kalman filter fusion.

[0121] S41, parameter initialization: Initialize the parameters of the adaptive Kalman filter, the specific parameters include: state transfer matrix , the observation matrix , initial process noise covariance , initial observation noise covariance , forgetting factor , initial state estimate , initial covariance .

[0122] S42, filtering process: for each time point in the first data sequence and the second data sequence , perform the following steps:

[0123] S43, prediction step: according to the state transfer matrix and the state estimate at the previous moment , calculate the prior state estimate at the current moment ; According to the state transfer matrix , the covariance of the previous moment and process noise covariance , calculate the prior covariance at the current moment .

[0124] S44, observation step: obtain the current data from the first data sequence and the second data sequence Satellite timing data after time denoising and removal of periodic terms and Roland timing data ,in Indicates taking The real part function of , expressed as:

[0125] .

[0126] S45, update step: according to the prior covariance and the observation matrix , calculate the Kalman gain ; According to the prior state estimate , Kalman gain and the observation vector , update the state estimate ; According to the Kalman gain and prior covariance , update the covariance .

[0127] S46, Adaptive adjustment step: Calculate observation residuals , according to the forgetting factor and observation residuals , adaptively adjust the observation noise covariance .

[0128] S47, fusion result output: the current State estimation at time i Stored as fused timing data Repeat S43 to S46 until all time points in the first data sequence and the second data sequence are traversed, and finally a fused timing data sequence is obtained. The fused timing data sequence is a complete one-dimensional timing data sequence. The final output is as follows: Figure 4 shown.

[0129] like Figure 6 As shown, this embodiment also provides a satellite and Loran timing data fusion system, including: a preprocessing module, configured to sequentially perform data truncation and mean removal on an acquired satellite timing data sequence to obtain a first mean-removed data sequence; sequentially perform data truncation and mean removal on the acquired Loran timing data sequence to obtain a second mean-removed data sequence; a decomposition and reconstruction module, configured to sequentially perform frequency domain conversion, singular value decomposition, and reconstruction on the first mean-removed data sequence and the second mean-removed data sequence to obtain a reconstructed frequency domain matrix; a periodic term removal module, configured to remove periodic term interference signals and perform time domain conversion on the reconstructed frequency domain matrix based on a set threshold in the frequency domain to obtain a first data sequence and a second data sequence, wherein the first data sequence is a data sequence after denoising the satellite timing data sequence and removing the periodic term, and the second data sequence is a data sequence after denoising the Loran timing data sequence and removing the periodic term; the set threshold is calculated based on the reconstructed frequency domain matrix; and a data fusion module, configured to perform adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filtering algorithm to obtain a fused timing data sequence.

[0130] The present invention also provides a satellite and Loran timing data fusion device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the satellite and Loran timing data fusion method when executing the computer program.

[0131] The present invention also provides a computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of the satellite and Loran timing data fusion method.

[0132] When the processor executes the computer program, the steps of fusing the above-mentioned satellite and Loran timing data are implemented, for example: performing data truncation and mean removal on the acquired satellite timing data sequence in sequence to obtain a first mean-removed data sequence; performing data truncation and mean removal on the acquired Loran timing data sequence in sequence to obtain a second mean-removed data sequence; performing frequency domain conversion, singular value decomposition, and reconstruction on the first mean-removed data sequence and the second mean-removed data sequence in sequence to obtain a reconstructed frequency domain matrix; based on a set threshold in the frequency domain, removing the periodic interference signal and performing time domain conversion on the reconstructed frequency domain matrix to obtain a first data sequence and a second data sequence, wherein the first data sequence is a data sequence after the satellite timing data sequence is denoised and the periodic term is removed, and the second data sequence is a data sequence after the Loran timing data sequence is denoised and the periodic term is removed; the set threshold is calculated based on the reconstructed frequency domain matrix; and an adaptive Kalman filtering algorithm is used to perform adaptive Kalman filtering fusion on the first data sequence and the second data sequence to obtain a fused timing data sequence.

[0133] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing preset functions, and the instruction segments are used to describe the execution process of the computer program in the satellite and Loran timing data fusion device. For example, the computer program can be divided into a preprocessing module, a decomposition and reconstruction module, a periodic term removal module, and a data fusion module; the specific functions of each module are as follows: the preprocessing module is used to sequentially perform data truncation and mean removal on the acquired satellite timing data sequence to obtain a first mean-removed data sequence; the acquired Loran timing data sequence is sequentially performed data truncation and mean removal to obtain a second mean-removed data sequence; the decomposition and reconstruction module is used to sequentially perform frequency domain conversion, singular value decomposition, and reconstruction on the first mean-removed data sequence and the second mean-removed data sequence to obtain a reconstructed frequency domain matrix; the periodic term removal module is used to remove the periodic term interference signal and perform time domain conversion on the reconstructed frequency domain matrix based on a set threshold in the frequency domain to obtain a first data sequence and a second data sequence, wherein the first data sequence is a data sequence after denoising the satellite timing data sequence and removing the periodic term, and the second data sequence is a data sequence after denoising the Loran timing data sequence and removing the periodic term; the set threshold is calculated based on the reconstructed frequency domain matrix; the data fusion module is used to perform adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filtering algorithm to obtain a fused timing data sequence.

[0134] The satellite and Loran timing data fusion device can be a computing device such as a desktop computer, laptop, PDA, or cloud server. The satellite and Loran timing data fusion device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the above examples of satellite and Loran timing data fusion devices do not limit the scope of satellite and Loran timing data fusion devices. The device may include more components than those described above, or a combination of certain components, or different components. For example, the satellite and Loran timing data fusion device may also include input and output devices, network access devices, buses, and the like.

[0135] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor. The processor serves as the control center for the satellite and Loran timing data fusion, connecting the entire satellite and various parts of the Loran timing data fusion equipment using various interfaces and lines.

[0136] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the satellite and Loran timing data fusion device by running or executing the computer program and / or module stored in the memory and calling the data stored in the memory.

[0137] The memory may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as sound playback or image playback); the data storage area may store data generated based on the use of the mobile phone (such as audio data and a phone book). Furthermore, the memory may include high-speed random access memory (RAM) and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0138] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the satellite and Loran timing data fusion method are implemented.

[0139] If the module / unit integrated with the satellite and Loran timing data fusion system is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium.

[0140] Based on this understanding, the present invention implements all or part of the processes in the above-mentioned satellite and Loran timing data fusion method, and can also be accomplished by using a computer program to instruct related hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of the above-mentioned satellite and Loran timing data fusion method. The computer program includes computer program code, which can be in source code form, object code form, executable file, or a pre-defined intermediate form.

[0141] The computer-readable storage medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0142] It should be noted that the content contained in the computer-readable storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable storage media do not include electrical carrier signals and telecommunication signals.

[0143] In summary, the present invention provides a method and related device for fusion of satellite and Loran timing data, which has the following advantages:

[0144] First, efficient noise reduction and periodic interference suppression: This invention innovatively uses frequency-domain SVD fusion denoising technology and a targeted periodic term removal strategy to achieve in-depth purification of satellite and Loran timing data. In frequency-domain SVD processing, based on the principle of singular value decomposition, the data is mapped to the frequency domain space. By screening and retaining the main singular values, the secondary components corresponding to the noise are effectively stripped away. Subsequently, a threshold is set to remove obvious periodic terms to further accurately eliminate periodic interference in the data. In this way, compared with traditional methods, the data purity is greatly improved, providing a high-quality data foundation for subsequent fusion and time synchronization calculations, and ensuring the reliability of timing data from the source.

[0145] Second, dynamic adaptive intelligent fusion: The introduction of an adaptive Kalman filter algorithm gives the system keen perception of dynamic data changes and intelligent adjustment capabilities. This method dynamically adjusts the observation noise covariance through the forgetting factor, adapting in real time to the characteristic differences between satellite and Loran timing data in different environments. In complex and changing signal environments, it can quickly respond to data fluctuations and dynamically optimize the fusion strategy, fully leveraging the high precision of satellite timing data and the strong anti-interference characteristics of Loran timing data to achieve deep fusion of the two data. Compared with fixed-parameter fusion algorithms, this significantly improves the accuracy and reliability of timing data, meeting the stringent requirements of high-precision time synchronization applications.

[0146] Third, strong universality and great potential for application expansion: The proposed method and related devices possess excellent versatility and scalability. Their design is independent of specific satellite or Loran timing system models and can seamlessly adapt to a variety of satellite timing systems (such as GPS, Beidou, etc.) combined with the Loran timing system. Furthermore, the core algorithm architecture and processing logic can be flexibly migrated to data fusion scenarios for other types of timing systems, effectively resolving the data fusion challenges between different timing systems. High-precision time synchronization can be stably achieved even in complex environments (such as areas of electromagnetic interference and satellite signal obstruction). This demonstrates high practical value and broad application prospects in numerous fields, including communications, navigation, power generation, and aerospace, and can promote the upgrading and development of time synchronization technologies in related industries.

[0147] The above embodiment is only one of the implementation methods that can realize the technical solution of the present invention. The scope of protection claimed by the present invention is not limited only to this embodiment, but also includes changes, replacements and other implementation methods that can be easily thought of by any technician familiar with this technical field within the technical scope disclosed by the present invention.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for fusing satellite and Loran timing data, characterized in that: include: performing data truncation and mean removal on the acquired satellite timing data sequence in sequence to obtain a first mean-removed data sequence; The obtained Loran timing data sequence is sequentially subjected to data truncation and mean removal to obtain a second mean-removed data sequence; Performing frequency domain conversion on the first de-meaned data sequence and the second de-meaned data sequence respectively to obtain a first frequency domain signal and a second frequency domain signal, combining the first frequency domain signal and the second frequency domain signal to obtain a frequency domain matrix, performing singular value decomposition on the frequency domain matrix and then reconstructing the matrix to obtain a reconstructed frequency domain matrix; Based on a set threshold in the frequency domain, the reconstructed frequency domain matrix is ​​subjected to periodic interference signal removal and time domain conversion to obtain a first data sequence and a second data sequence; the first data sequence is a satellite timing data sequence after denoising and removing periodic terms, and the second data sequence is a Loran timing data sequence after denoising and removing periodic terms; the set threshold is calculated based on the reconstructed frequency domain matrix; An adaptive Kalman filter algorithm is used to perform adaptive Kalman filter fusion on the first data sequence and the second data sequence to obtain a fused timing data sequence.

2. The method for fusion of satellite and Loran timing data according to claim 1, characterized in that: The acquired satellite timing data sequence is sequentially subjected to data truncation and mean removal to obtain a first mean-removed data sequence; The obtained Loran timing data sequence is sequentially subjected to data truncation and mean removal to obtain a second mean-removed data sequence, including: respectively intercepting valid data segments of the satellite timing data sequence and the Loran timing data sequence to obtain a first original data sequence and a second original data sequence, and making the first original data sequence and the second original data sequence have the same length; The first original data sequence and the second original data sequence are subjected to mean removal processing respectively to obtain a first mean-removed data sequence and a second mean-removed data sequence. The specific formula is as follows: In the formula, dataG1 represents the first mean-removed data sequence; dataLY1 represents the second mean-removed data sequence; dataG represents the first original data sequence; dataLY represents the second original data sequence; mean(dataG) represents the mean of the first original data sequence; mean(dataLY) represents the mean of the second original data sequence.

3. The method for fusion of satellite and Loran timing data according to claim 1, characterized in that: The first de-meaned data sequence and the second de-meaned data sequence are respectively subjected to frequency domain conversion to obtain a first frequency domain signal and a second frequency domain signal, the first frequency domain signal and the second frequency domain signal are combined to obtain a frequency domain matrix, and the frequency domain matrix is ​​subjected to singular value decomposition and then reconstructed to obtain a reconstructed frequency domain matrix, including: Perform column vector conversion on the first de-meaned data sequence and the second de-meaned data sequence respectively to obtain a first column vector and a second column vector; Performing frequency domain conversion on the first column vector and the second column vector respectively using fast Fourier transform to obtain a first frequency domain signal and a second frequency domain signal; Combining the first frequency domain signal and the second frequency domain signal to obtain a frequency domain matrix; Performing singular value decomposition on the frequency domain matrix to obtain a singular value matrix; the singular value matrix includes a left singular vector matrix, a singular value diagonal matrix and a right singular vector matrix; Perform singular value screening on the singular value diagonal matrix to construct a new singular value diagonal matrix; The frequency domain matrix is ​​reconstructed using the left singular vector matrix, the new singular value diagonal matrix and the right singular vector matrix to obtain a reconstructed frequency domain matrix.

4. The method for fusion of satellite and Loran timing data according to claim 1, characterized in that: Before removing the periodic interference signal and performing time domain conversion on the reconstructed frequency domain matrix based on the set threshold in the frequency domain to obtain the first data sequence and the second data sequence, the method includes: According to the maximum absolute value of the elements in the reconstructed frequency domain matrix, the set threshold is calculated and the formula is: threshold=0.01×max(|M denoised (:)|) In the formula, threshold represents the set threshold, M denoised Represents the reconstructed frequency domain matrix; max(|M denoised (:)|) represents the maximum absolute value of the elements in the reconstructed frequency domain matrix.

5. The method for fusion of satellite and Loran timing data according to claim 1, characterized in that: The method of removing periodic interference signals and performing time domain conversion on the reconstructed frequency domain matrix based on a set threshold in the frequency domain to obtain a first data sequence and a second data sequence includes: The elements in the reconstructed frequency domain matrix whose absolute values ​​are greater than the set threshold are set to 0 to remove the periodic interference signal from the reconstructed frequency domain matrix; By using inverse fast Fourier transform, the first row and the second row of the reconstructed frequency domain matrix with the periodic interference signal removed are respectively converted into the time domain to obtain a first data sequence and a second data sequence.

6. The method for fusion of satellite and Loran timing data according to claim 1, characterized in that: The method of performing adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filtering algorithm to obtain a fused timing data sequence includes: Step 1: Obtain first data and second data at the current moment from the first data sequence and the second data sequence, and combine the first data and the second data into an observation vector; the first data is the satellite timing data after denoising and removing the periodic term, and the second data is the Loran timing data after denoising and removing the periodic term; Step 2: Calculate the Kalman gain based on the pre-calculated prior covariance and the preset observation matrix; update the state estimate based on the pre-calculated prior state estimate, the Kalman gain, and the observation vector; output the updated state estimate as the fused timing data; Step 3: Adaptively adjust the observation noise covariance based on the preset forgetting factor and the pre-calculated observation residual; Steps 1, 2, and 3 are executed cyclically until the data corresponding to all moments in the first data sequence and the second data sequence are traversed, and the fused timing data corresponding to each moment is obtained to obtain a fused timing data sequence.

7. The method for fusion of satellite and Loran timing data according to claim 6, characterized in that: The step 1 includes: Initialize the parameters of the adaptive Kalman filter; the parameters include the state transfer matrix, the observation matrix, the process noise covariance, the observation noise covariance, the forgetting factor, the state estimate and covariance; Before the step 2, the method includes: The prior state estimate at the current moment is calculated based on the state transfer matrix and the state estimate at the previous moment; the prior covariance at the current moment is calculated based on the state transfer matrix, the covariance at the previous moment and the process noise covariance.

8. A satellite and Loran timing data fusion system, characterized in that: include: a preprocessing module, configured to sequentially perform data truncation and mean removal on the acquired satellite timing data sequence to obtain a first mean-removed data sequence; The obtained Loran timing data sequence is sequentially subjected to data truncation and mean removal to obtain a second mean-removed data sequence; a decomposition and reconstruction module, configured to perform frequency domain conversion on the first de-meaned data sequence and the second de-meaned data sequence, respectively, to obtain a first frequency domain signal and a second frequency domain signal, combine the first frequency domain signal and the second frequency domain signal to obtain a frequency domain matrix, perform singular value decomposition on the frequency domain matrix, and then reconstruct the frequency domain matrix to obtain a reconstructed frequency domain matrix; a periodic term removal module configured to remove periodic term interference signals and perform time domain conversion on the reconstructed frequency domain matrix based on a set threshold in the frequency domain to obtain a first data sequence and a second data sequence; the first data sequence is a satellite timing data sequence denoised and periodic term removed, and the second data sequence is a Loran timing data sequence denoised and periodic term removed; the set threshold is calculated based on the reconstructed frequency domain matrix; The data fusion module is used to perform adaptive Kalman filtering fusion on the first data sequence and the second data sequence using an adaptive Kalman filtering algorithm to obtain a fused timing data sequence.

9. A satellite and Loran timing data fusion device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the satellite and Loran timing data fusion method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it is used to implement the steps of the satellite and Loran timing data fusion method according to any one of claims 1 to 7.

Images