Enhanced beidou precise timing method based on receiver clock model

Abstract

This invention relates to the field of radio direction finding technology, specifically to an enhanced BeiDou precise time synchronization method based on a receiver clock model. The method includes: acquiring multi-source data and preprocessing it; establishing a BeiDou receiver clock model, optimizing model parameters using the least squares method, outputting a first clock drift value and a first coefficient, and using a machine learning model to predict a second clock drift value and a second coefficient; then, using a Bayesian fusion model to combine the BeiDou receiver clock model and the machine learning model to output a predicted clock drift value and calculate a predicted time offset value; calculating the carrier-to-noise ratio loss value, autocorrelation function distortion value, and phase change value to generate signal quality indicators, and using a weighted allocation mechanism to calculate an error compensation value; finally, correcting the clock deviation based on the error compensation value. This invention improves the accuracy of clock drift prediction and effectively corrects clock deviation by combining multiple models, thereby improving the accuracy of time synchronization in mountainous areas.

Description

Technical Field

[0001] This invention relates to the field of radio direction finding technology, specifically to an enhanced BeiDou precision timing method based on a receiver clock model. Background Technology

[0002] The BeiDou Navigation Satellite System, independently developed by my country, is widely used in communication, timing, and positioning. It holds significant strategic value, particularly in remote mountainous areas and complex terrain where communication conditions are limited. However, mountainous terrain is complex, often characterized by mountain obstructions and multipath interference, affecting the reliability of BeiDou signals and reducing or even interrupting timing accuracy. Furthermore, the complex electromagnetic environment in mountainous areas, with weak signals and prolonged coverage, increases positioning and timing errors, posing challenges to critical applications such as emergency communication and disaster monitoring.

[0003] Existing BeiDou satellite timing technology can provide high-precision timing services, with advantages such as low cost, wide coverage, and good continuity, especially demonstrating good stability and reliability in open areas and conventional scenarios. However, its adaptability in mountainous scenarios remains insufficient. First, these technologies have relatively simple modeling of local clock drift in receiver clocks, which may not be able to cope with the dynamic changes in clock errors in complex mountainous environments, making it difficult to maintain stable timing accuracy. Second, observation errors caused by multipath interference and signal obstruction are large, and these error correction models are mainly based on fixed parameters, lacking real-time dynamic adjustment capabilities, and cannot effectively cope with changes in the mountainous environment. Third, mountainous communication scenarios have high real-time requirements for timing equipment, but most of these systems rely on offline processing and post-processing data, which is difficult to meet the demand for precise BeiDou satellite timing in real-time communication in mountainous areas. Therefore, the accuracy of BeiDou satellite timing needs to be further improved.

[0004] To address this, an enhanced BeiDou precision time synchronization method based on a receiver clock model is proposed. Summary of the Invention

[0005] The purpose of this invention is to provide an enhanced BeiDou precise timing method based on a receiver clock model. This method involves acquiring and preprocessing multi-source data; establishing a BeiDou receiver clock model; optimizing model parameters using the least squares method; outputting a first clock drift value and a first coefficient; and using a machine learning model to predict a second clock drift value and a second coefficient. Then, a Bayesian fusion model is used to combine the BeiDou receiver clock model and the machine learning model to output a predicted clock drift value and calculate a predicted time offset value. The carrier-to-noise ratio loss, autocorrelation function distortion, and phase change values ​​are calculated to generate signal quality indicators, and an error compensation value is calculated using a weighted allocation mechanism. Finally, the clock deviation is corrected based on the error compensation value. By combining multiple models, the accuracy of clock drift prediction is improved, and clock deviation is effectively corrected, thereby enhancing the precision of BeiDou satellite timing in mountainous areas.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] An enhanced BeiDou precision time synchronization method based on a receiver clock model includes:

[0008] Step S10: Acquire multi-source data, including environmental data, GIS data, meteorological data, signal data, and historical clock data; preprocess the multi-source data to obtain processed data;

[0009] Step S20: Establish a BeiDou receiver clock model, which is used to calculate the clock drift caused by the processed data; optimize the parameters of the BeiDou receiver clock model using the least squares method based on the processed data, and output a first clock drift value and a first coefficient; use a machine learning model on the processed data to output a second clock drift value and a second coefficient.

[0010] Step S30: Based on the first clock drift value, the first coefficient, the second clock drift value, and the second coefficient, output a clock drift prediction value using a Bayesian fusion model; calculate a time offset prediction value based on the clock drift prediction value;

[0011] Step S40: Calculate the carrier-to-noise ratio loss, autocorrelation function distortion, and phase change value for the processed data to generate a signal quality index; calculate the error compensation value using a weighting mechanism based on the signal quality index and the predicted time offset.

[0012] Step S50: Correct the clock deviation according to the error compensation value.

[0013] Furthermore, the multi-source data includes:

[0014] The environmental data includes altitude and receiver data;

[0015] The GIS data includes topographic features, geological structure, and vegetation cover;

[0016] The meteorological data includes rainfall, wind speed, temperature, humidity, and air pressure;

[0017] The signal data includes signal power, noise power, direct path signal, and multipath signal;

[0018] The clock history data includes historical drift information of the clock.

[0019] Furthermore, the process of establishing the BeiDou receiver clock model includes:

[0020] A temperature influence model is established for the processed data, and the temperature frequency drift is output; the temperature influence model is a linear model of the crystal oscillator frequency changing with temperature;

[0021] A vibration influence model is established for the processed data, and the vibration frequency drift is output; the vibration influence model is a model of the crystal oscillator frequency changing with the vibration intensity;

[0022] A power fluctuation impact model is established on the processed data, and the power frequency drift is output; the power fluctuation impact model is a model of how the crystal oscillator frequency changes with power fluctuations;

[0023] The clock model of the Beidou receiver is obtained by summing the temperature frequency drift, the vibration frequency drift, the power frequency drift, and the random error.

[0024] Furthermore, the implementation process of the least squares method includes:

[0025] The processed data is acquired, and a design matrix is ​​constructed based on the temperature sequence, vibration intensity sequence, and vibration intensity sequence, as shown below:

[0026]

[0027] Where X is the design matrix, [T1,T2…T n [V1, V2…V] represents a temperature sequence, where T0 is the reference temperature. n [ΔPS1, ΔPS2…ΔPS] represents the vibration intensity sequence. n [] represents the power fluctuation amplitude sequence, where 1, 2, and n are sequence numbers;

[0028] Based on the design matrix, and according to the BeiDou receiver clock model, minimizing the sum of squared errors is expressed as:

[0029] S=(yX×k) t ·(yX×k);

[0030] Where S is the least squares objective value, y is the actual clock drift vector, X is the design matrix, k is the parameter vector, and t is the transpose sign;

[0031] Taking the derivative of S with respect to k and setting the derivative to zero, the parameter vector k is output as follows:

[0032] k=(X T ×X) -1 ×X T ×y;

[0033] Among them, X T ×X is the covariance matrix of the design matrix, X T ×y is the inner product of the design matrix and the actual clock drift vector.

[0034] Furthermore, the implementation process of the Bayesian fusion model includes:

[0035] Define a prior distribution, which is a normal distribution generated based on the first clock drift value and the first coefficient;

[0036] Define a likelihood function, which is the normal distribution generated based on the second clock drift value and the second coefficient;

[0037] According to Bayes' theorem, the prior distribution is combined with the likelihood function to generate the posterior distribution;

[0038] The mean of the posterior distribution is used as the final clock drift prediction.

[0039] Furthermore, the calculation process of the signal quality index includes:

[0040] Divide the signal power by the noise power to obtain the received signal carrier-to-noise ratio; obtain the ideal carrier-to-noise ratio to obtain the carrier-to-noise ratio loss value;

[0041] Calculate the autocorrelation function of the received signal and the autocorrelation function of the ideal signal to obtain the distortion value of the autocorrelation function;

[0042] The instantaneous phase of the received signal is extracted using the Hilbert transform to obtain the phase change value;

[0043] The signal quality index is obtained by adding the carrier-to-noise ratio loss value, the autocorrelation function distortion value, and the phase change value.

[0044] Furthermore, the weight allocation mechanism includes:

[0045] The signal quality index is mapped to a signal error, and the error compensation value is obtained by weighted summing the error with the time offset prediction value.

[0046] If the change in the signal quality index exceeds a preset threshold, the gradient descent algorithm is used to adjust the signal quality weight and the time offset prediction weight; otherwise, no adjustment is made.

[0047] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0048] 1. This invention employs a BeiDou receiver clock model to model the linear impact of environmental and equipment factors such as temperature, vibration, and power fluctuations on crystal oscillator frequency drift in complex mountainous environments. It then optimizes parameters using the least squares method to accurately capture linear clock drift in mountainous areas. Simultaneously, a machine learning model is introduced to perform nonlinear predictions of complex clock drift relationships using multi-source data. This comprehensively models the complex nonlinear factors in mountainous environments, effectively predicting nonlinear clock drift caused by environmental disturbances, thereby enhancing the stability and accuracy of BeiDou precision time synchronization.

[0049] 2. This invention, based on Bayesian theory, integrates the output of the BeiDou receiver clock model with the output of a machine learning model. By leveraging the advantages of both model-driven and data-driven approaches, it reduces the bias that might be introduced by a single model prediction in complex signal environments, such as mountainous areas, through probabilistic inference. Next, a prior distribution and a likelihood function are defined, integrating information from both sources into a posterior distribution. The mean of this posterior distribution is then used as the final clock drift prediction value. This significantly improves the accuracy of clock drift prediction, especially in situations with high uncertainty or missing data, thereby enhancing the stability and precision of BeiDou's precise time synchronization.

[0050] 3. This invention comprehensively reflects the signal transmission quality in mountainous areas by defining signal quality indicators that take into account factors such as carrier-to-noise ratio loss, autocorrelation function distortion, and phase change. Next, a weighted allocation mechanism is introduced to calculate the error compensation value by weighted summation of the signal quality indicators and time offset prediction values. The weights are dynamically adjusted according to the magnitude of signal quality changes, effectively addressing signal quality fluctuations in mountainous communications, rationally allocating the influence weights of signal quality and time offset, reducing reliance on low-quality signals, and thus improving timing accuracy and adaptability in complex signal environments. Attached Figure Description

[0051] Figure 1 This invention provides a flowchart illustrating an enhanced BeiDou precision time synchronization method based on a BeiDou receiver clock model.

[0052] Figure 2 This invention provides a schematic flowchart for constructing a BeiDou receiver clock model. Detailed Implementation

[0053] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0054] Please see Figures 1 to 2 This invention provides an enhanced BeiDou precise timing method based on a receiver clock model, the technical solution of which is as follows:

[0055] Example 1 is as follows:

[0056] An enhanced BeiDou precision time synchronization method based on a receiver clock model includes:

[0057] Step S10: Acquire multi-source data, including environmental data, GIS data, meteorological data, signal data, and historical clock data; preprocess the multi-source data to obtain processed data.

[0058] Furthermore, the multi-source data includes:

[0059] The environmental data includes altitude and receiver data;

[0060] The GIS data includes topographic features, geological structure, and vegetation cover;

[0061] The meteorological data includes rainfall, wind speed, temperature, humidity, and air pressure;

[0062] The signal data includes signal power, noise power, direct path signal, and multipath signal;

[0063] The clock history data includes historical drift information of the clock.

[0064] Specifically, mountainous elevation data in environmental data affects signal propagation paths and attenuation, while receiver operating temperature and voltage parameters influence clock stability. GIS data affects signal propagation paths and multipath effects; understanding terrain helps optimize signal processing. Meteorological data impacts receiver performance and clock stability, and signal data directly affects signal quality and clock synchronization accuracy, requiring real-time monitoring to optimize error compensation. Historical clock data is used to identify long-term trends and periodic changes in clock drift, facilitating subsequent analysis and modeling. Acquiring this multi-source data enables higher accuracy in the entire clock synchronization and error compensation process, making it suitable for complex and variable mountainous communication environments, thus meeting the requirements for high-precision clock synchronization and stable communication using BeiDou.

[0065] Step S20: Establish a BeiDou receiver clock model, which is used to calculate the clock drift caused by the processed data; optimize the parameters of the BeiDou receiver clock model using the least squares method based on the processed data, and output a first clock drift value and a first coefficient; use a machine learning model on the processed data to output a second clock drift value and a second coefficient.

[0066] Furthermore, the process of establishing the BeiDou receiver clock model includes:

[0067] A temperature influence model is established for the processed data, and the temperature frequency drift is output. The temperature influence model is a linear model of the crystal oscillator frequency changing with temperature, expressed as:

[0068] Δf T =α T ×(T-T0);

[0069] Where, Δf T For temperature frequency drift, α T T is the temperature coefficient, where T is the current temperature and T0 is the reference temperature.

[0070] If the temperature range in the mountainous area is 5 to 25 degrees Celsius, the reference temperature can be the median of this range, which is 15 degrees Celsius, or the reference temperature recommended by the crystal oscillator manufacturer can be used.

[0071] A vibration influence model is established for the processed data, and the vibration frequency drift is output. The vibration influence model is a model of the crystal oscillator frequency changing with the vibration intensity, expressed as:

[0072] Δf V =α V ×V;

[0073] Where, Δf V For the frequency shift of the vibration, α V V is the vibration coefficient, and V is the vibration intensity.

[0074] A power fluctuation impact model is established for the processed data, and the power frequency drift is output. The power fluctuation impact model is a model of how the crystal oscillator frequency changes with power fluctuations, expressed as:

[0075] Δf PS =α PS ×ΔPS;

[0076] Where, Δf PS For power frequency drift, α PS ΔPS represents the power supply fluctuation coefficient, and ΔPS represents the power supply fluctuation amplitude.

[0077] Summing the temperature frequency drift, the vibration frequency drift, the power frequency drift, and the random error yields the BeiDou receiver clock model, expressed as:

[0078] Δf=Δf T +Δf V +Δf PS +∈;

[0079] Where Δf is the clock drift, Δf T For temperature frequency drift, Δf V For the frequency shift of the vibration, Δf PS For power frequency drift, ∈ represents random error caused by other unconsidered factors, which can be set to -10. -8 Up to 10 -8 Small fluctuations within a few seconds.

[0080] Specifically, in mountainous areas and other regions with significant temperature differences, the crystal oscillator frequency of the receiver is particularly affected by temperature changes, making temperature drift a key factor in clock drift. By establishing a temperature influence model, the impact of temperature on frequency can be quantified, thereby accurately calculating clock drift caused by changes in ambient temperature. Next, vibration intensity refers to the acceleration of vibration in the environment where the crystal oscillator is located. Vibration affects the physical performance of the crystal oscillator. By introducing a vibration intensity influence model, frequency drift caused by vibration can be effectively estimated, thus improving clock accuracy. Furthermore, power supply fluctuations can lead to instability in the crystal oscillator's power supply, affecting frequency output. By establishing a power supply fluctuation influence model, the frequency drift characteristics of the crystal oscillator can be captured more comprehensively. Additionally, there are some uncontrollable or unmodeled random errors (such as environmental noise and system noise). By introducing random error terms, the BeiDou receiver clock model can be better adapted to real-time conditions in mountainous areas.

[0081] By quantitatively modeling temperature, vibration, power fluctuations, and random errors, and constructing a comprehensive BeiDou receiver clock model, the accuracy of calculations is not only improved in theory, but also a reliable data foundation is laid for the accurate prediction and effective compensation of clock drift.

[0082] Furthermore, the implementation process of the least squares method includes:

[0083] Acquire sufficient processing data, including temperature, reference temperature, power supply fluctuations, vibration intensity, and actual clock drift value, organize it into a format suitable for regression analysis, and construct a design matrix, represented as follows:

[0084]

[0085] Where X is the design matrix, [T1,T2…T n [V1, V2…V] represents a temperature sequence, where T0 is the reference temperature.n [ΔPS1, ΔPS2…ΔPS] represents the vibration intensity sequence. n [] represents the power fluctuation amplitude sequence, where 1, 2, and n are sequence numbers;

[0086] The sum of squared errors minimized according to the BeiDou receiver clock model is expressed as:

[0087] S=(yX×k) t ·(yX×k);

[0088] Where S is the least squares objective value, y is the actual clock drift vector, X is the design matrix, and k is the parameter vector and k = [α]. T ,α V …α PS ], α T α is the temperature coefficient. V Let α be the vibration coefficient. PS Here, t represents the power fluctuation coefficient, and t is the transpose sign.

[0089] Taking the derivative of S with respect to k and setting the derivative to zero, the parameter vector k is output as follows:

[0090] k=(X T ×X) -1 ×X T ×y;

[0091] Among them, X T ×X is the covariance matrix of the design matrix, X T ×y is the inner product of the design matrix and the actual clock drift vector.

[0092] The least squares method, by acquiring multi-dimensional data such as temperature, reference temperature, power fluctuations, and vibration intensity, constructs a design matrix that comprehensively reflects environmental factors. This matrix accurately captures the impact of these factors on clock drift, providing a strong basis for precise time synchronization in mountainous areas. Furthermore, the least squares method can adjust model parameters under changing mountainous environments, enabling the BeiDou receiver clock model to adapt to environmental changes and thus ensuring the accuracy of the receiver clock.

[0093] Next, the clock drift output by the optimized BeiDou receiver clock model is subtracted from the actual clock drift to obtain the residual. Then, based on the number of parameters and the total number of data points, the degrees of freedom are calculated. The standard deviation of the residuals is then calculated based on the degrees of freedom and the residuals to obtain the first coefficient, which represents the overall uncertainty of the BeiDou receiver clock model and indicates the typical deviation between the model's predicted values ​​and the observed values. This coefficient is expressed as:

[0094]

[0095] Where σ1 is the first coefficient, e iLet n be the i-th residual, n be the total number of data points, m be the number of parameters, and nm be the degrees of freedom.

[0096] Furthermore, the machine learning model can choose to use a random forest model to predict clock drift in mountainous areas. The mean and standard deviation of all decision tree predictions in the random forest are calculated. The mean serves as the model's second clock drift value, and the standard deviation serves as a quantification of the uncertainty of the prediction results, reflecting the range of fluctuation in the predicted values. Next, using the normal distribution assumption, the 95% prediction interval is calculated, thus effectively quantifying the uncertainty of the prediction results and providing more comprehensive information. Finally, the average of all standard deviations is taken to obtain the second coefficient.

[0097] Step S30: Based on the first clock drift value, the first coefficient, the second clock drift value, and the second coefficient, output a clock drift prediction value using a Bayesian fusion model; calculate a time offset prediction value based on the clock drift prediction value.

[0098] Furthermore, the implementation process of the Bayesian fusion model includes:

[0099] Define a prior distribution, which is a normal distribution generated based on the first clock drift value and the first coefficient, expressed as:

[0100]

[0101] Where θ is the target variable in Bayesian inference, i.e., the final predicted clock drift value, P(θ) is the prior distribution, y1 is the first clock drift value, and σ1 is the first coefficient. For y1 and It follows a normal distribution.

[0102] Define a likelihood function, which is a normal distribution generated based on the second clock drift value and the second coefficient, and express it as follows:

[0103]

[0104] in, The clock offset value obtained from experimental measurements can also be equal to σ². Let y2 be the likelihood function, y2 be the second clock drift value, and σ2 be the second coefficient. For y2 and It follows a normal distribution.

[0105] According to Bayes' theorem, the prior distribution is combined with the likelihood function to generate the posterior distribution;

[0106] The mean of the posterior distribution is used as the final clock drift prediction.

[0107] Specifically, in mountainous environments with complex and non-linear changes, machine learning models can accurately capture the unique characteristics of mountainous areas, while the BeiDou receiver clock model can calculate the overall trend. Through Bayesian fusion, the determinism of the first clock drift value is combined with the flexibility of the second clock drift value. When new influencing factors appear in the mountainous environment, causing changes in clock drift, the model can quickly adapt and adjust its prediction strategy, automatically optimizing model weights. Thus, in the variable mountainous environment, not only is the accuracy of clock drift prediction improved, but the robustness of the model is also enhanced.

[0108] Step S40: Calculate the carrier-to-noise ratio loss, autocorrelation function distortion, and phase change value for the processed data to generate a signal quality index; calculate the error compensation value using a weighted allocation mechanism based on the signal quality index and the predicted time offset value.

[0109] Furthermore, the calculation process of the signal quality index includes:

[0110] Divide the signal power by the noise power to obtain the received signal carrier-to-noise ratio (CNR); obtain the ideal CNR, and subtract the ideal CNR from the received signal CNR to obtain the CNR loss value.

[0111] The autocorrelation function of the received signal and the autocorrelation function of the ideal signal are calculated to obtain the distortion value of the autocorrelation function, which is expressed as:

[0112]

[0113] ΔR(τ)=R y (τ)-R s (τ);

[0114] Where y(t) is the signal captured by the receiver, s(t) is the direct path signal, and a i τ is the amplitude factor of the multipath signal. i Let be the delay of the multipath signal, i be the index of the multipath signal, N be the total number of multipath signals, t be the time variable, and R be the time delay. y (τ) is the autocorrelation function of the received signal, R s (τ) is the autocorrelation function of an ideal signal, with a peak value at τ = 0, ∞ being the infinite sign, dt being the time derivative, and ΔR(τ) being the distortion value of the autocorrelation function.

[0115] The instantaneous phase of the received signal is extracted using the Hilbert transform to obtain the phase change value;

[0116] The signal quality index is obtained by adding the carrier-to-noise ratio loss value, the autocorrelation function distortion value, and the phase change value.

[0117] Specifically, the carrier-to-noise ratio (CNR) loss is used to quantitatively assess the impact of environmental noise and interference on signal quality, thereby helping to optimize the receiver's noise immunity. The autocorrelation function distortion (ACF) value can quantitatively describe the impact of multipath interference on the signal's timing characteristics, and the phase change value is used to detect the signal's phase stability, providing a basis for subsequent phase compensation and correction. By combining the CNR loss, ACF distortion, and phase change values, the overall signal quality is quantitatively assessed, providing a comprehensive signal quality evaluation standard. This facilitates dynamic optimization of subsequent parameters, thereby ensuring the accuracy of BeiDou timing in mountainous areas.

[0118] Furthermore, the weight allocation mechanism includes:

[0119] The signal quality index is mapped to a signal error, and the error compensation value is obtained by weighted summing the error with the time offset prediction value.

[0120] If the change in the signal quality index exceeds a preset threshold, the gradient descent algorithm is used to adjust the signal quality weight and the time offset prediction weight; otherwise, no adjustment is made.

[0121] Specifically, signal quality indicators reflect the current signal condition, encompassing carrier-to-noise ratio loss, autocorrelation function distortion, and phase changes, all of which directly affect the accuracy of clock synchronization. Time offset predictions, derived from the BeiDou receiver clock model and historical data, reveal the long-term trend and expectations of clock drift. By using a weighted summation method, real-time signal quality and model-predicted time offsets can be integrated to comprehensively assess the required error compensation. For example, in environments with significant signal quality fluctuations, the weight of signal quality indicators can be increased; conversely, in scenarios with stable signal quality, the time offset predictions can be relied upon more heavily. Flexibly adjusting the weighting of each indicator adapts to the diverse performance requirements of mountainous areas, thereby improving the accuracy of BeiDou time synchronization in mountainous regions.

[0122] Step S50: Correct the clock deviation according to the error compensation value.

[0123] In summary, this invention utilizes both a BeiDou receiver clock model and a machine learning model to calculate clock drift values ​​and corresponding coefficients, fully leveraging the strengths of different models to improve the accuracy and stability of clock drift prediction. In mountainous environments with complex signal propagation paths and high uncertainty, Bayesian fusion of prediction results from different models comprehensively considers the uncertainty and reliability of each model, further enhancing prediction accuracy. Furthermore, by combining real-time signal quality and time offset predictions and dynamically adjusting weights, this invention ensures that clock synchronization can respond and adjust rapidly in mountainous environments with significant signal quality fluctuations, adapting to different communication needs and changes in mountainous areas, thereby improving the accuracy of BeiDou time synchronization.

[0124] Example 2 is as follows:

[0125] Mountainous area A has complex terrain with irregularly distributed valleys and ridges, making signals susceptible to obstruction. Furthermore, it experiences significant climate variations and strong seasonality, potentially causing receiver fluctuations. To ensure the high real-time performance of BeiDou time synchronization in mountainous areas such as emergency rescue and time synchronization with mountain communication base stations, based on Example 1, an enhanced BeiDou precise time synchronization method based on a receiver clock model is presented, including:

[0126] Figure 1 This invention provides a flowchart of an enhanced BeiDou precision timing method based on a receiver clock model.

[0127] refer to Figure 1 Step S10: Acquire multi-source data, including environmental data, GIS data, meteorological data, signal data, and historical clock data; preprocess the multi-source data to obtain processed data;

[0128] refer to Figure 1 In step S20: Establish a BeiDou receiver clock model, which is used to calculate the clock drift caused by the processed data; optimize the parameters of the BeiDou receiver clock model using the least squares method based on the processed data, and output a first clock drift value and a first coefficient; use a machine learning model on the processed data to output a second clock drift value and a second coefficient.

[0129] refer to Figure 1 In step S30: Based on the first clock drift value, the first coefficient, the second clock drift value, and the second coefficient, output a clock drift prediction value using a Bayesian fusion model; calculate a time offset prediction value based on the clock drift prediction value;

[0130] refer to Figure 1 In step S40: Calculate the carrier-to-noise ratio loss, autocorrelation function distortion, and phase change value for the processed data to generate a signal quality index; calculate the error compensation value using a weighting allocation mechanism based on the signal quality index and the predicted time offset.

[0131] refer to Figure 1 Step S50: Correct the clock deviation according to the error compensation value.

[0132] Furthermore, the multi-source data includes:

[0133] The environmental data includes altitude and receiver data;

[0134] The GIS data includes topographic features, geological structure, and vegetation cover;

[0135] The meteorological data includes rainfall, wind speed, temperature, humidity, and air pressure;

[0136] The signal data includes signal power, noise power, direct path signal, and multipath signal;

[0137] The clock history data includes historical drift information of the clock.

[0138] Specifically, in this embodiment, the altitude and GIS data of each receiver within the A mountainous area are obtained using an existing GIS database and updated monthly. Real-time operating status data is obtained through internal temperature sensors and power status monitoring, and updated every 10 seconds. Local weather stations near the communication base station are used to monitor meteorological data in real time, and updated every 10 minutes. Signal power and noise power are measured in real time using a built-in power meter in the receiver, and updated every 10 seconds. Historical clock data is obtained from the clock synchronization log and updated every 12 hours. This data is then cleaned and unified according to the minimum time step to obtain processed data.

[0139] Figure 2 This invention provides a schematic flowchart for constructing a BeiDou receiver clock model.

[0140] Furthermore, such as Figure 2 As shown, the process of establishing the BeiDou receiver clock model includes:

[0141] A temperature influence model is established for the processed data, and the temperature frequency drift is output. The temperature influence model is a linear model of the crystal oscillator frequency changing with temperature.

[0142] A vibration influence model is established for the processed data, and the vibration frequency drift is output. The vibration influence model is a model of how the crystal oscillator frequency changes with the vibration intensity.

[0143] A power fluctuation impact model is established on the processed data, and the power frequency drift is output. The power fluctuation impact model is a model of how the crystal oscillator frequency changes with power fluctuations.

[0144] The clock model of the Beidou receiver is obtained by summing the temperature frequency drift, the vibration frequency drift, the power frequency drift, and the random error.

[0145] Furthermore, the implementation process of the least squares method includes:

[0146] The processed data is obtained, and a design matrix is ​​constructed, represented as follows:

[0147]

[0148] Where X is the design matrix, [T1,T2…T n[V1, V2…V] represents a temperature sequence, where T0 is the reference temperature. n [ΔPS1, ΔPS2…ΔPS] represents the vibration intensity sequence. n [] represents the power fluctuation amplitude sequence, where 1, 2, and n are sequence numbers;

[0149] The sum of squared errors minimized according to the BeiDou receiver clock model is expressed as:

[0150] S=(yX×k) t ·(yX×k);

[0151] Where S is the least squares objective value, y is the actual clock drift vector, X is the design matrix, and k is the parameter vector k = [α] T ,α V …α PS ], where t is the transpose symbol;

[0152] Taking the derivative of S with respect to k and setting the derivative to zero, the parameter vector k is output as follows:

[0153] k=(X T ×X) -1 ×X T ×y;

[0154] Among them, X T ×X is the covariance matrix of the design matrix, X T ×y is the inner product of the design matrix and the actual clock drift vector.

[0155] Table 1 Example of Input Data

[0156] Serial Number temperature Reference temperature Power fluctuations Vibration intensity Actual clock drift value 1 30 25 5 2 1.2 2 35 25 10 4 2.4 3 25 25 5 1 1.1 4 40 25 15 3 3.2

[0157] Table 2 Example of Output Data

[0158] Parameter name Temperature coefficient Power fluctuation coefficient Vibration intensity coefficient estimated value 0.1 0.1 0.25

[0159] Table 3 Examples of Machine Learning Model Output Results

[0160] Serial Number Forecast Mean Predicted standard deviation 95% prediction interval lower bound 95% prediction interval upper limit 1 0.12 0.02 0.08 0.16 2 0.15 0.03 0.09 0.21 3 0.10 0.01 0.08 0.12

[0161] Specifically, sufficient historical clock data is obtained, as shown in Table 1, including temperature, reference temperature, power supply fluctuations, vibration intensity, and actual clock drift value. This data is then formatted for regression analysis. The optimized parameters of the BeiDou receiver clock model are obtained using the least squares method, as shown in Table 2. For example, a temperature coefficient of 0.1 means that for every 1°C increase in temperature compared to the reference temperature, the clock drift value increases by 0.1 units, thus improving the calculation accuracy of the BeiDou receiver clock model. If the model's accuracy does not meet the requirements, the least squares method is used again to output updated parameters.

[0162] Furthermore, a random forest model is used to output a second clock drift value and a second coefficient on the processed data. The output results are shown in Table 3. The predicted value of each sample has a 95% probability of falling within the prediction interval, indicating that the reliability of each prediction result is relatively high. Next, the mean of all prediction standard deviations is taken, and the second coefficient is 0.02.

[0163] Furthermore, the implementation process of the Bayesian fusion model includes:

[0164] Define a prior distribution, which is a normal distribution generated based on the first clock drift value and the first coefficient;

[0165] Define a likelihood function, which is the normal distribution generated based on the second clock drift value and the second coefficient;

[0166] According to Bayes' theorem, the prior distribution is combined with the likelihood function to generate the posterior distribution;

[0167] The mean of the posterior distribution is used as the final clock drift prediction.

[0168] Specifically, to demonstrate the effectiveness of the Bayesian fusion model, the random forest model and the BeiDou receiver clock model before fusion, as well as the SVM model in machine learning, were compared. The prediction results are shown in Table 4. It can be seen that the Bayesian fusion model outperforms the single BeiDou receiver clock model, random forest model, and SVM model in terms of mean absolute error and root mean square error, thus improving the prediction accuracy.

[0169] Table 4 Model Prediction Results

[0170] Model type Mean Absolute Error Root mean square error Beidou receiver clock model 0.18 0.20 Random Forest Model 0.15 0.17 Support Vector Machine Model 0.16 0.18 Bayesian fusion model 0.13 0.14

[0171] Next, the time offset prediction value is calculated based on the clock drift prediction value, and is expressed as:

[0172]

[0173] Where DEV(Δt) is the predicted time offset, Δf is the predicted clock drift, t0 is the start time, t is the current time, and dt is the derivative with respect to time.

[0174] Furthermore, the calculation process of the signal quality index includes:

[0175] Divide the signal power by the noise power to obtain the received signal carrier-to-noise ratio; obtain the ideal carrier-to-noise ratio to obtain the carrier-to-noise ratio loss value;

[0176] Calculate the autocorrelation function of the received signal and the autocorrelation function of the ideal signal to obtain the distortion value of the autocorrelation function;

[0177] The instantaneous phase of the received signal is extracted using the Hilbert transform to obtain the phase change value;

[0178] The signal quality index is obtained by adding the carrier-to-noise ratio loss value, the autocorrelation function distortion value, and the phase change value.

[0179] Furthermore, the weight allocation mechanism includes:

[0180] The signal quality index is mapped to a signal error, and the error compensation value is obtained by weighted summing the error with the time offset prediction value.

[0181] If the change in the signal quality index exceeds a preset threshold, the gradient descent algorithm is used to adjust the signal quality weight and the time offset prediction weight; otherwise, no adjustment is made.

[0182] Specifically, a machine learning model, such as an LSTM model, is used to map the signal quality metric to a possible time error range, i.e., to predict the current signal error. During the initial run, initial weights are set for the signal quality metric and the predicted time offset, ensuring that their sum is 1. In this embodiment, the initial signal quality weight W... q,old Weight W of time offset predicted value t,old The values ​​are set to 0.5 and 0.5 respectively. Next, the magnitude of the change in the signal quality index, ΔQ, is calculated and compared with a preset threshold of 0.05 to determine whether the weights need to be adjusted. If the magnitude of the change is greater than 0.05, the weights are adjusted; otherwise, the weights remain unchanged.

[0183] Define the loss function as follows:

[0184] L=(O current -E previous ) 2 ;

[0185] Where L is the loss value, O current E represents the currently measured time offset value. previous This is the error compensation value calculated at the previous moment.

[0186] Next, the updated signal quality weight W is calculated using the gradient descent algorithm. q,new Weight W of time offset predicted value t,new As shown in Table 5, an example of the weight allocation mechanism is given. By combining the signal quality index with the predicted clock offset and using a weighted summation method to calculate the error compensation value, the real-time signal condition and long-term clock offset trend can be comprehensively considered. This ensures a rapid response when the signal quality changes significantly, optimizes the error compensation effect, and improves the accuracy of BeiDou time synchronization in mountainous communication environments.

[0187] Table 5. Example of the weight allocation mechanism process

[0188] Time step ΔQ L <![CDATA[W q,old ]]> <![CDATA[W t,old ]]> <![CDATA[W q,new ]]> <![CDATA[W t,new ]]> 1 - - 0.5 0.5 0.5 0.5 2 0.02 0.0004 0.5 0.5 0.5 0.5 3 0.04 0.0001 0.5 0.5 0.5 0.5 4 0.14 0.0289 0.5 0.5 0.5005 0.4995 5 0.07 0.01 0.5005 0.4995 0.50106 0.49894

[0189] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An enhanced BeiDou precision time synchronization method based on a receiver clock model, characterized in that, include: Acquire multi-source data, including environmental data, GIS data, meteorological data, signal data, and historical clock data; The multi-source data is preprocessed to obtain processed data; A BeiDou receiver clock model is established, and the process of establishing the receiver clock model includes: A temperature influence model is established for the processed data, and the temperature frequency drift is output; the temperature influence model is a linear model of the crystal oscillator frequency changing with temperature; A vibration influence model is established for the processed data, and the vibration frequency drift is output; the vibration influence model is a model of the crystal oscillator frequency changing with the vibration intensity; A power fluctuation impact model is established on the processed data, and the power frequency drift is output; the power fluctuation impact model is a model of how the crystal oscillator frequency changes with power fluctuations; The temperature frequency drift, vibration frequency drift, power frequency drift, and random error are summed to obtain the BeiDou receiver clock model; the BeiDou receiver clock model is used to calculate the clock drift caused by the processed data; the parameters of the BeiDou receiver clock model are optimized using the least squares method based on the processed data, and a first clock drift value and a first coefficient are output; a machine learning model is used on the processed data to output a second clock drift value and a second coefficient; Based on the first clock drift value, the first coefficient, the second clock drift value, and the second coefficient, a Bayesian fusion model is used to output a predicted clock drift value; the implementation process of the Bayesian fusion model includes: Define a prior distribution, which is a normal distribution generated based on the first clock drift value and the first coefficient; Define a likelihood function, which is the normal distribution generated based on the second clock drift value and the second coefficient; According to Bayes' theorem, the prior distribution is combined with the likelihood function to generate the posterior distribution; The mean of the posterior distribution is used as the final clock drift prediction; the time offset prediction is calculated based on the clock drift prediction. The carrier-to-noise ratio loss, autocorrelation function distortion, and phase change are calculated on the processed data to generate a signal quality index; based on the signal quality index and the predicted time offset, an error compensation value is calculated using a weighting mechanism. The clock deviation is corrected based on the error compensation value.

2. The enhanced BeiDou precision timing method based on a receiver clock model according to claim 1, characterized in that, The multi-source data includes: The environmental data includes altitude and receiver data; The GIS data includes topographic features, geological structure, and vegetation cover; The meteorological data includes rainfall, wind speed, temperature, humidity, and air pressure; The signal data includes signal power, noise power, direct path signal, and multipath signal; The clock history data includes historical drift information of the clock.

3. The enhanced BeiDou precise timing method based on a receiver clock model according to claim 1, characterized in that, The implementation process of the least squares method includes: The processed data is acquired, and a design matrix is ​​constructed based on the temperature sequence, vibration intensity sequence, and vibration intensity sequence. Based on the design matrix, the sum of squared errors of the BeiDou receiver clock model is minimized to obtain the least squares target value. The parameter vector is output by taking the derivative of the least squares objective value with respect to the parameter vector and setting the derivative to zero.

4. The enhanced BeiDou precision timing method based on a receiver clock model according to claim 1, characterized in that, The calculation process for the signal quality index includes: Divide the signal power by the noise power to obtain the received signal carrier-to-noise ratio; obtain the ideal carrier-to-noise ratio to obtain the carrier-to-noise ratio loss value; Calculate the autocorrelation function of the received signal and the autocorrelation function of the ideal signal to obtain the distortion value of the autocorrelation function; The instantaneous phase of the received signal is extracted using the Hilbert transform to obtain the phase change value; The signal quality index is obtained by adding the carrier-to-noise ratio loss value, the autocorrelation function distortion value, and the phase change value.

5. The enhanced BeiDou precise timing method based on a receiver clock model according to claim 1, characterized in that, The weight allocation mechanism includes: The signal quality index is mapped to a signal error, and the error compensation value is obtained by weighted summing the error with the time offset prediction value. If the change in the signal quality index exceeds a preset threshold, the gradient descent algorithm is used to adjust the signal quality weight and the time offset prediction weight; otherwise, no adjustment is made.

Images