GNSS (Global Navigation Satellite System) double-antenna direction-finding adaptive satellite screening algorithm

Abstract

The invention provides a GNSS (Global Navigation Satellite System) double-antenna direction-finding adaptive satellite screening algorithm, which relates to the technical field of satellite navigation, and comprises the following steps of: completing the first round of visible satellite screening by initializing an elevating angle and a carrier-to-noise ratio threshold, and then constructing an environment state vector according to the number of satellites, an average carrier-to-noise ratio, a carrier-to-noise ratio standard deviation and an average elevating angle; and the adaptive function is driven to reset the threshold in real time, so that the satellite availability in a complex scene is ensured. The multi-dimensional comprehensive scoring function further eliminates low-quality satellites, and the course precision evaluation function selects an optimal subset according to geometric intensity and signal quality. According to the layered process, the robustness of the system to environment change is remarkably enhanced, the course precision and stability are improved, the calculation load is reduced, and the system can always stably, continuously and precisely give the course in a complex environment.

Description

Technical Field

[0001] The present invention relates to the field of satellite navigation technology, and in particular to a GNSS dual-antenna direction finding adaptive satellite screening algorithm. Background Art

[0002] Dual-antenna direction-finding technology based on satellite navigation signals is widely used on vehicles, aircraft, and ships to obtain precise heading information in real time. The basic principle is to deploy two GNSS (Global Navigation Satellite System) receiving antennas with a known baseline length on the vehicle. By measuring the carrier phase difference between the same navigation satellite signal reaching the two antennas in real time, interferometry is used to calculate the direction of incidence of the satellite signal relative to the baseline, thereby determining the vehicle's heading angle in the local horizontal coordinate system.

[0003] To ensure direction finding accuracy and reliability, existing dual-antenna direction finding systems typically use a fixed satellite screening strategy. This involves presetting a set of empirical thresholds during system initialization, including: Static satellite elevation angle threshold (usually ≥15°~20°), used to eliminate satellites with low elevation angles and significant multipath effects; Static C / N threshold (usually ≥35 dB-Hz) to eliminate satellites with weak signal power or interference; Minimum number of available satellites (usually ≥4) to ensure the geometric strength of the solution.

[0004] However, the above-mentioned fixed threshold scheme has exposed the following shortcomings in actual complex environments: in open scenes, the number of visible satellites is sufficient (≥8), but the static satellite altitude angle threshold and static carrier-to-noise ratio threshold may exclude some satellites with slightly lower signal quality but favorable geometric distribution, resulting in reduced redundancy of observation information and limited accuracy; in scenes with severe obstruction such as urban canyons, boulevards or overpasses, the static satellite altitude angle threshold and static carrier-to-noise ratio threshold will greatly reduce the number of available satellites, even falling below the minimum solution requirement (<4), resulting in the system being unable to output a valid heading.

[0005] Therefore, there is an urgent need for a satellite screening algorithm that can adaptively adjust the screening strategy according to the real-time satellite environment and the dynamic state of the carrier, so as to balance the contradiction between "satellite quality" and "satellite quantity" and improve the continuity and accuracy of the dual-antenna direction finding system in complex scenarios. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention aims to propose a GNSS dual-antenna direction finding adaptive satellite screening algorithm to solve the problems mentioned in the above background technology section.

[0007] The present invention is achieved through the following technical solutions: A GNSS dual-antenna direction finding adaptive satellite screening algorithm includes the following steps: S1. Initialize the altitude angle threshold and the carrier-to-noise ratio threshold, and based on the initial values ​​of the altitude angle threshold and the carrier-to-noise ratio threshold, screen out all satellites that meet the visibility conditions; S2. Based on all satellites that meet the visibility conditions, construct an environmental state vector of the dual-antenna direction finding system at time t. The environmental state vector includes the number of satellites that meet the visibility conditions, the average carrier-to-noise ratio of the satellites, the standard deviation of the satellite carrier-to-noise ratio, and the average satellite elevation angle; S3. Build an adaptive function that dynamically adjusts the altitude angle threshold and the carrier-to-noise ratio threshold according to the environmental state vector, and re-screens candidate satellites that meet the threshold requirements based on the adjusted altitude angle threshold and the carrier-to-noise ratio threshold; S4. Construct a multi-dimensional satellite comprehensive scoring function, evaluate the quality of all candidate satellites in step S3 using the comprehensive scoring function, and eliminate satellites with too low scores; S5. The remaining satellites in step S4 are screened out to obtain the optimal satellite set through the heading accuracy evaluation function.

[0008] The beneficial effects of the present invention are as follows: the present invention completes the first round of visible satellite screening by initializing the altitude angle and carrier-to-noise ratio thresholds, and then constructs the environmental state vector based on the number of satellites, average carrier-to-noise ratio, carrier-to-noise ratio standard deviation and average altitude angle, driving the adaptive function to reset the threshold in real time to ensure satellite availability in complex scenarios. The multi-dimensional comprehensive scoring function further eliminates low-quality satellites, and the heading accuracy evaluation function selects the optimal subset based on geometric strength and signal quality. This hierarchical process significantly enhances the system's robustness to environmental changes, improves heading accuracy and stability, reduces computational load, and enables the system to always provide heading stably, continuously and with high precision in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 This is a flow chart of a GNSS dual-antenna direction finding adaptive satellite screening algorithm according to the present invention. DETAILED DESCRIPTION

[0010] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. It should be noted that the description of these embodiments is intended to help understand the present invention, but does not constitute a limitation of the present invention. In addition, the technical features involved in the various embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.

[0011] Reference Figure 1 As shown, the present invention provides a GNSS dual-antenna direction finding adaptive satellite screening algorithm, comprising the following steps: S1. Initialize the altitude angle threshold and the carrier-to-noise ratio threshold, and based on the initial values ​​of the altitude angle threshold and the carrier-to-noise ratio threshold, screen out all satellites that meet the visibility conditions; S2. Based on all satellites that meet the visibility conditions, construct an environmental state vector of the dual-antenna direction finding system at time t. The environmental state vector includes the number of satellites that meet the visibility conditions, the average carrier-to-noise ratio of the satellites, the standard deviation of the satellite carrier-to-noise ratio, and the average satellite elevation angle; S3. Build an adaptive function that dynamically adjusts the altitude angle threshold and the carrier-to-noise ratio threshold according to the environmental state vector, and re-screens candidate satellites that meet the threshold requirements based on the adjusted altitude angle threshold and the carrier-to-noise ratio threshold; S4. Construct a multi-dimensional satellite comprehensive scoring function, evaluate the quality of all candidate satellites in step S3 using the comprehensive scoring function, and eliminate satellites with too low scores; S5. The remaining satellites in step S4 are screened out to obtain the optimal satellite set through the heading accuracy evaluation function.

[0012] As a beneficial effect of the above embodiment: This algorithm dynamically adjusts the altitude angle threshold and carrier-to-noise ratio threshold based on the environmental state vector, effectively balancing the contradiction between "satellite quality" and "satellite quantity." When a large number of satellites meet the visibility conditions (open scenarios), the altitude angle and carrier-to-noise ratio thresholds can be lowered to include more satellites with slightly lower signal quality but favorable geometric distribution in the observation, increasing the redundancy of the observation information and improving the solution accuracy. When a large number of satellites meet the visibility conditions (occlusion scenarios), the thresholds can be dynamically lowered and the minimum number of satellites can be flexibly adjusted to avoid system failure due to a sudden drop in the number of satellites and ensure direction finding continuity. Compared to the rigid strategy of fixed thresholds, this function achieves dynamic matching of screening parameters with the real-time environment.

[0013] Using a comprehensive scoring function to assess the quality of individual satellites across multiple dimensions provides a more comprehensive reflection of their actual availability. This avoids the misjudgment of satellites with slightly lower quality but favorable geometry due to a single threshold, or the inadvertent inclusion of satellites that meet performance standards but are actually experiencing interference. This improves the accuracy of satellite quality assessments and provides a reliable basis for subsequent satellite portfolio optimization.

[0014] The optimal satellite set is selected using a heading accuracy evaluation function. When the number of satellites is sufficient, satellite combinations with a more optimal geometric distribution are prioritized to improve heading solution accuracy. When the number of satellites is limited, combinations that maximize heading accuracy while ensuring basic quality are selected, avoiding the blind abandonment of valid solutions due to insufficient satellites. This step further unlocks the value of satellite observation information, ensuring that the system can maintain continuity and maximize accuracy in complex scenarios.

[0015] As a preferred embodiment, the algorithm further includes step S5: S6. Perform double-difference ambiguity resolution optimization on the satellite set selected in step S5, narrow the ambiguity search space based on baseline length and condition number constraints, and use the enhanced LAMBDA algorithm to solve it. Calculate the heading angle based on the solution result and output it.

[0016] As a beneficial effect of the above-described embodiment, from the perspective of ambiguity search space optimization, baseline length is an inherent parameter of a dual-antenna system. Its constraint can significantly reduce the range of possible ambiguity values. In complex scenarios, satellite signals are susceptible to occlusion, reflection, and other factors, and ambiguity candidate values ​​often contain a large amount of redundancy. The baseline length constraint can physically limit the reasonable range of ambiguity, avoiding ineffective searches. Furthermore, the condition number reflects the rationality of the satellite geometric distribution. The smaller the condition number, the higher the geometric strength. Based on this constraint, unreasonable ambiguity candidate values ​​caused by poor geometric distribution can be further eliminated, focusing the search space on high-probability solutions. This significantly reduces computational complexity and improves solution efficiency. This is particularly true in scenarios with a small number of satellites or complex geometric distributions (such as urban canyons). This avoids the time-consuming or solution-failure problems of traditional algorithms caused by excessively large search spaces.

[0017] Calculating and outputting heading angles based on optimized solution results directly improves heading angle accuracy and stability. In open scenes, optimized ambiguity resolution reduces heading deviations caused by accumulated observation noise. In heavily obscured scenes, even with a limited number of available satellites, ambiguity resolution effectiveness is maintained, ensuring the continuity of heading angle output by narrowing the search space and enhancing the algorithm's fault tolerance.

[0018] As a preferred embodiment, the algorithm further includes step S7: S7. Establish a dynamic parameter update mechanism to re-evaluate the environmental status and adjust relevant parameters when significant changes in the environment are detected.

[0019] As a beneficial effect of the above-described embodiment, this mechanism can detect sudden changes in the environment in real time, ensuring that system parameters are aligned with the current scenario. In practical applications, a vehicle may rapidly move from an open landscape into an urban canyon (e.g., a vehicle entering a tunnel entrance) or from a tree-lined avenue onto an open road. These dramatic environmental changes can lead to sudden changes in the number of visible satellites, signal quality, and geometric distribution. The dynamic parameter update mechanism monitors these changes in real time (e.g., a sudden drop in the number of satellites, a significant fluctuation in the carrier-to-noise ratio, etc.) and triggers a reassessment of the environmental state, thus avoiding the lag inherent in the traditional fixed-parameter model, where the number of visible satellites drops from eight to three. For example, if a vehicle moves from an open area into a densely populated area, and the number of visible satellites drops from eight to three, the mechanism quickly identifies this significant change, reassesses the environmental state vector, and adjusts parameters such as the altitude angle threshold, carrier-to-noise ratio threshold, and minimum number of satellites to ensure that the system retains as many available satellites as possible in the new environment and maintains the continuity of heading output.

[0020] In step S1, the visibility determination function , used to determine whether a single satellite at time t meets the visibility condition; Visibility determination function as follows: if ,but

[0021] ,but ; In the above formula, Indicates that at time t the satellite Altitude angle; represents the altitude mask, Indicates that at time t the satellite carrier-to-noise ratio; Indicates the carrier-to-noise ratio mask.

[0022] Among them, at the moment t equals 0, dB-Hz, Indicates the initial value of the carrier-to-noise ratio threshold, , Indicates the initial value of the altitude angle threshold.

[0023] In step S2, the environment state vector is as follows:

[0024] represents the environmental state vector of the dual-antenna direction finding system at time t, where the parameters in the environmental state vector are calculated using the following formula: , represents the number of satellites that meet the visibility conditions at time t; , It represents the average carrier-to-noise ratio of all satellites that meet the visibility conditions at time t, which is used to reflect the overall strength level of the current satellite signal; , It represents the standard deviation of the carrier-to-noise ratio of all satellites that meet the visibility conditions at time t, and is used to reflect the dispersion of the carrier-to-noise ratio of each satellite; , It represents the average elevation angle of all satellites that meet the visibility conditions at time t, and is used to reflect the overall elevation angle distribution of satellites in the sky.

[0025] In step S3, the adaptive function includes an altitude angle threshold adaptive function, and the altitude angle threshold adaptive function is as follows: ; In the above formula, Indicates the final altitude angle threshold for satellite screening, its value depends on the number of currently visible satellites ; Indicates the initial value of the altitude angle threshold; Indicates the adaptive adjustment amount of the altitude angle threshold.

[0026] Different intervals are divided by setting the number of satellites that meet the visibility conditions, including neutral interval, increasing interval and decreasing interval; When the number of satellites that meet the visibility conditions is within the neutral range, ; When the number of satellites that meet the visibility conditions increases, ; When the number of satellites meeting the visibility conditions is in the decreasing range, ; As a preferred embodiment, in this algorithm, , and provide adaptive adjustment as follows: When the number of satellites that meet the visibility conditions is: hour, ; When the number of satellites that meet the visibility conditions is: hour, ; When the number of satellites that meet the visibility conditions is: hour, ; When the number of satellites that meet the visibility conditions is: hour, ; When the number of satellites that meet the visibility conditions is: hour, ; And Set a lower bound constraint: .

[0027] As a beneficial effect of the above embodiment: according to the number of visible satellites in the environmental state vector Divide into multiple intervals and set the adaptive adjustment amount in a targeted manner. ), increase the altitude angle threshold ( ), select high-quality satellites with high elevation angles and small multipath effects to improve solution accuracy; in occlusion scenarios (such as ), significantly reducing the altitude angle threshold ( ), incorporate more medium and low elevation angle satellites to avoid system failure due to insufficient number, and allow the elevation angle threshold to accurately adapt to different satellite resource abundance scenarios.

[0028] The altitude threshold adaptive function adjusts the altitude threshold value based on the number of satellites that meet the visibility conditions. This design directly addresses the core contradiction between the number of satellites and the altitude threshold in different scenarios. In open scenes, when the number of visible satellites is sufficient, the altitude threshold adaptive function can appropriately increase the altitude threshold, while ensuring satellite quality, while avoiding the introduction of multipath interference due to the inclusion of too many low-elevation-angle satellites. In scenes with severe obstruction, when the number of visible satellites is small, the altitude threshold adaptive function can actively lower the altitude threshold, including more medium- and low-elevation-angle satellites that might have been excluded by the fixed threshold, effectively alleviating the problem of insufficient satellites and ensuring the continuity of the system in complex environments. This method, based on the number of satellites as the core adjustment basis, achieves a dynamic matching of the altitude threshold and the abundance of satellite resources, balancing the relationship between "quality" and "quantity."

[0029] Setting a lower limit constraint can prevent the elevation angle threshold from being lowered without limit. Even if there are very few satellites, satellites with ultra-low elevation angles and severe multipath can be filtered out, reducing invalid observation interference, ensuring the reliability of carrier phase difference measurement, providing a stable data basis for heading angle calculation, and improving system output stability.

[0030] In step S3, the adaptive function includes a carrier-to-noise ratio threshold adaptive function, and the carrier-to-noise ratio threshold adaptive function is as follows: ; In the above formula, Indicates the carrier-to-noise ratio threshold ultimately used for satellite screening; Indicates the initial value of the carrier-to-noise ratio threshold; The weight of the impact of the number of satellites on the adjustment of the carrier-to-noise ratio threshold; represents the satellite number adjustment function; Indicates the impact weight of environmental quality; represents the environmental quality adjustment function; ; In the above formula, represents the hyperbolic tangent function; Indicates the number of satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the number of satellite reference points, Used to define the "reference basis for changes in the number of satellites" and to make the function revolve around Symmetrical adjustment ( Quantity above / below When the number of satellites changes, the adjustment changes in the opposite direction); A represents the gain factor, which is used to control the "amplification factor" of the hyperbolic tangent function output and determines the "intensity of the impact of changes in the number of satellites on the adjustment amount"; B represents the scaling factor, which is used to control the "sensitivity" of the function to changes in the number of satellites; C represents the offset constant, which is used to adjust the overall "offset" of the function to ensure that the baseline value of the adjustment amount adapts to the basic logic of the carrier-to-noise ratio threshold;

[0031] In the above formula, D represents the gain factor; Indicates the environmental quality benchmark point;

[0032] In the above formula, 、 Respectively represent the weights of environmental quality dimensions; Represents the average carrier-to-noise ratio of all satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the normalized maximum value of the carrier-to-noise ratio; It represents the standard deviation of the carrier-to-noise ratio of all satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the normalized maximum value of the standard deviation of the carrier-to-noise ratio; Represents the average elevation angle of all satellites that meet the visibility conditions, obtained through the environmental state vector.

[0033] As a preferred embodiment, dB-Hz, and the parameters in the above formula are quantified as follows: ; ; Weight coefficient: = ; = ; ; right Set up constraints: [dB-Hz]; As a beneficial effect of the above embodiment: The environmental quality assessment is constructed by comprehensively considering the average carrier-to-noise ratio of all satellites that meet the visibility conditions, the standard deviation of the carrier-to-noise ratio of satellites that meet the visibility conditions, and the average elevation angle of all satellites that meet the visibility conditions, covering signal strength, stability, geometric distribution dimensions, and satellite quantity adjustment function. By combining the number of satellites that meet visibility conditions, the carrier-to-noise ratio threshold can simultaneously respond to dual environmental changes in "number of satellites" and "signal quality + geometric distribution", accurately adapting to scenarios such as open (many satellites, good signals) and blocked (few satellites, poor signals), solving the problem that fixed thresholds cannot take into account different environments.

[0034] The parameters in the formula (such as A, B, C, 、 With clearly defined quantification (e.g., [e.g., [e.g., quantization parameters]), engineers can debug the algorithm based on the carrier (vehicle / aircraft) and scenario requirements, facilitating its deployment in real-world devices. Compared to abstract strategies, quantized parameters make threshold adjustment logic clearer and more reproducible, reducing engineering implementation complexity and improving algorithm practicality and scalability.

[0035] Noise ratio standard deviation is included in environmental quality assessment ( ), can identify environments with large signal fluctuations (multiple interferences, multipath), and by adjusting the carrier-to-noise ratio threshold, remove severely interfered satellites, enhancing the system's anti-interference capabilities. In complex scenarios such as urban canyons, it effectively reduces the impact of noisy satellites on the solution, improving the robustness of the heading output.

[0036] This function represents the final carrier-to-noise ratio threshold used for satellite screening. This is the output of the algorithm's dynamic adjustments and directly determines whether a satellite is filtered due to a substandard carrier-to-noise ratio. This function serves as a key metric for satellite screening, balancing the key parameters of "retaining high-quality signals" and "accommodating the number of satellites to the environment."

[0037] is the “influence weight” of the number of satellites on the adjustment of the carrier-to-noise ratio threshold; The "influence weight" of environmental quality on the carrier-to-noise ratio threshold adjustment. These two factors are used to prioritize "number of satellites" versus "environmental quality" in adjusting the carrier-to-noise ratio threshold, allowing for flexible adaptation to different scenarios (e.g., prioritizing quantity or quality).

[0038] The adaptive CNR threshold function is calculated based on the number of satellites meeting visibility conditions, the environmental quality adjustment function, and the base CNR threshold value, embodying a refined adjustment logic that combines both quantity and quality. When a sufficient number of visible satellites are available, the environmental quality adjustment function assesses overall signal quality based on the average CNR, CNR standard deviation, and average elevation angle. If overall signal quality is high, the CNR threshold can be appropriately raised to select higher-quality satellites and reduce redundant information. If some satellites have slightly lower signal quality but favorable geometric distribution, the adaptive CNR threshold function will flexibly relax the threshold to retain these satellites that contribute to improved accuracy. In obstruction scenarios with a small number of satellites, the adaptive CNR threshold function can appropriately lower the threshold based on the environmental quality adjustment function's assessment while ensuring basic signal quality, avoiding further reductions in the number of available satellites due to overly stringent standards. Furthermore, the introduction of a base CNR threshold value provides a benchmark for adjustments, ensuring that the threshold does not fluctuate indefinitely, balancing stability and adaptability. This multi-parameter linkage adjustment method not only avoids the rigidity of the fixed carrier-to-noise ratio threshold in complex environments, but also dynamically balances the strictness of screening according to the overall situation of signal quality, further improving the rationality of satellite screening.

[0039] As a preferred embodiment, in step S4, the comprehensive scoring function of a single satellite is as follows: ; In the above formula, It indicates that each satellite is given a comprehensive score by weighted summing up the scores of six dimensions: carrier-to-noise ratio, elevation angle, azimuth diversity, carrier phase quality, multipath effect, and double-difference residual; Represents the weight vector of each dimension; Indicates the Satellite No. Ratings of dimensions.

[0040] As a preferred embodiment, in step S3, the comprehensive scoring function of a single satellite is as follows: ; In the above formula, It indicates that each satellite is given a comprehensive score by weighted summing up the scores of six dimensions: carrier-to-noise ratio, elevation angle, azimuth diversity, carrier phase quality, multipath effect and double difference residual; Represents the weight vector of each dimension; Indicates the Satellite No. Ratings of dimensions.

[0041] As a beneficial effect of the above embodiment, by abandoning the one-sidedness of single-dimensional evaluation, the satellite's value to direction finding is comprehensively characterized from six dimensions: carrier-to-noise ratio (reflecting signal strength and anti-interference performance), elevation angle (related to multipath effects and geometric distribution), azimuth angle diversity (affecting the geometric strength of the solution), carrier phase quality (determining phase measurement accuracy), multipath effects (direct interference signal authenticity), and double-difference residual (reflecting the satellite's adaptability in the combined solution). This comprehensive coverage of signal quality, spatial distribution, and solution coordination makes satellite quality assessment more accurate and comprehensive, preventing high-quality satellites from being incorrectly screened out due to weaknesses in a single dimension, or low-quality satellites from being incorrectly selected due to strengths in a particular dimension.

[0042] Specifically, the sub-functions in the comprehensive scoring function are as follows: The carrier-to-noise ratio scoring function is: ,in, Indicates that at time t The carrier-to-noise ratio of the satellite; The carrier-to-noise ratio scoring function maps the carrier-to-noise ratio to the [0,1] range, accurately distinguishing between strong and weak signals. A 30 dB-Hz threshold is used, and a 25 dB-Hz scaling interval is used. This clearly quantifies the importance of signal quality for direction finding and avoids incorrect satellite selection due to ambiguous carrier-to-noise ratio judgments.

[0043] The altitude angle scoring function is: ,in, Indicates that at time t The altitude angle of the satellite; Through the altitude angle scoring function, with π / 18 as the low threshold and π / 3 as the high threshold, the impact of altitude angle on multipath effect is quantified, and satellites with high elevation angle and weak multipath are given priority to improve the solution accuracy.

[0044] The azimuthal diversity scoring function is: ,in, Indicates that at time t Azimuth of satellites: ,in, represents the azimuth of all satellites that meet the visibility conditions at time t; It means that at time t, all satellites that meet the visibility conditions are summed up to obtain the total azimuth angle. represents all satellites that meet the visibility conditions at time t; The azimuth diversity scoring function quantifies the diversity of the azimuth distribution by comparing the difference between the satellite azimuths and the mean. The greater the difference, the higher the score, which guides the selection of dispersed satellites, enhances the geometric strength of the solution, and addresses the accuracy shortcomings caused by satellite clustering.

[0045] The carrier phase quality scoring function is: ,in, Indicates that at time t Phase observation noise of satellites; represents the standard deviation of the double-difference carrier phase noise at time t; The carrier phase quality scoring function accurately measures the reliability of phase measurements based on the ratio of phase observation noise to the standard deviation of double-difference noise. Lower noise levels correspond to higher scores, ensuring phase difference calculation accuracy and laying a solid foundation for carrier phase differential direction finding.

[0046] The multipath effect scoring function is: ,in, represents the multipath interference intensity of the i-th satellite at time t; , represents the pseudorange observation value of the i-th satellite at time t, represents the geometric distance of the i-th satellite at time t, represents the speed of light; Through the multipath effect scoring function, the deviation between pseudorange and geometric distance ( Quantify multipath interference. The smaller the deviation, the higher the score, effectively identifying and avoiding satellites with severe multipath, reducing their interference with direction finding results.

[0047] The double-difference residual score function is: ,in, represents the absolute value of the double-difference observation residual of the i-th satellite at time t, represents the a priori standard deviation of the double-difference residual; The double-difference residual scoring function quantifies the satellite's suitability for the combined solution by measuring the ratio of the absolute value of the double-difference observation residual to the residual standard deviation. Smaller residuals are associated with higher scores, allowing for the selection of satellites with good coordination and improving overall solution stability.

[0048] Weight vector of each dimension ; In the above embodiment, the inputs to each sub-function (such as carrier-to-noise ratio, elevation angle, and azimuth angle) are all real-time satellite status, enabling dynamic scoring based on the environment. In urban canyons, where azimuth diversity and elevation angle scores are weighted highly, satellites that can break through obstructions are prioritized. In open areas, where carrier-to-noise ratio and carrier phase quality are weighted highly, satellites with pure signals are prioritized. This allows the quality assessment to be tailored to the scenario and enhances the system's environmental adaptability. Multi-dimensional scoring can identify satellite weaknesses under varying interference conditions, proactively eliminating poorly performing satellites during screening and reducing interference sources. Even with sudden environmental changes (such as sudden multipath), the scoring system can still select relatively high-quality satellites, maintaining stable system output and enhancing direction-finding robustness.

[0049] Score in the carrier-to-noise ratio dimension: This dimension reflects the strength and degree of interference of satellite signals. This dimension's score can screen out satellites with stable signal power, providing a reliable signal basis for direction finding and avoiding direction finding errors caused by weak signals or strong interference. The elevation angle score considers the satellite's elevation angle. High-elevation-angle satellites are less susceptible to multipath effects. Low-elevation-angle satellites, while potentially more susceptible to multipath, can be helpful in geometric distribution in certain scenarios. A comprehensive score provides a reasonable balance and avoids the one-sidedness of a single elevation angle threshold. Scoring for Azimuth Diversity: In direction finding, the diversity of the satellite azimuth distribution significantly impacts solution accuracy. A concentration of satellites in a single azimuth reduces geometric strength. Considering this dimension in the overall scoring can help select satellites with a more even distribution, improving the geometric accuracy of the solution. Scoring in the carrier phase quality dimension: This dimension is directly related to the accuracy of phase measurement. Scoring in this dimension can eliminate satellites with large phase jitter and high noise, ensuring the reliability of phase observations and laying the foundation for accurate calculation of carrier phase differences. Scoring for the multipath effect dimension: This is a significant factor influencing satellite signal quality. Its evaluation is included in the comprehensive scoring, effectively identifying and reducing the weight of satellites with strong multipath signals, reducing the interference of multipath errors on direction-finding results, and improving the system's anti-interference capability. The double-difference residual can reflect the actual performance of the satellite in the combined solution. By scoring this dimension, satellites with small residuals and good coordination with other satellites during the solution process can be screened out, thereby improving the solution consistency and stability of the overall satellite combination. In summary, this six-dimensional comprehensive scoring function breaks through the limitations of traditional single- or limited-dimensional evaluations, comprehensively assessing satellite quality across multiple dimensions, including signal quality, geometric distribution, and solution performance. It accurately identifies high-quality satellites while avoiding misjudging satellites with slightly lower signal quality that may be beneficial for improving direction-finding accuracy. This provides a more scientific and comprehensive basis for subsequent satellite combination optimization and selection, thereby improving the accuracy and reliability of dual-antenna direction-finding systems in complex scenarios.

[0050] First, satellites with too low scores are eliminated through the satellite comprehensive scoring function, and the remaining satellites are screened out to obtain the optimal satellite set through the heading accuracy evaluation function; The heading accuracy evaluation function is as follows: ; In the above formula, represents the selected satellite set; represents the heading angle variance; represents the weight coefficient; Represents a penalty term.

[0051] Among them, the penalty The function is as follows:

[0052] In the above formula, Indicates the penalty weight for the deviation in the number of satellites; Indicates the total number of target selected satellites; Indicates the number of satellites in the actually selected satellite combination S; The penalty weight representing the deviation of the system distribution; Indicates the number of targets for each satellite system; Indicates the number of satellites in each system in the actually selected satellite combination S.

[0053] For example: If the target is 8 and the actual number is 6, then , the quantity deviation penalty is 4 If the target GPS is 3 and BeiDou is 3, the actual GPS is 2 and BeiDou is 4. , the system distribution penalty is 2 .

[0054] As a beneficial effect of the above embodiment: , directly related to the impact of the satellite set S on the heading solution accuracy in the form of heading angle variance. A smaller variance indicates better coordination of the satellite combination in terms of geometric distribution and signal quality, leading to more accurate heading solutions. This provides a quantitative basis for selecting high-quality satellite sets and avoids the blind practice of selecting satellites based on experience.

[0055] pass , penalizing satellite ensemble "shortcomings" (such as insufficient satellite numbers or extreme distribution). When the number of satellites is too small, the penalty increases, guiding the system to balance the number and distribution of satellites while pursuing accuracy. This addresses issues where simply pursuing low variance leads to infeasible satellite ensembles (such as insufficient satellite numbers or poor interference immunity), achieving a balance between "accuracy and constraints," avoiding the selection of too many satellites (high computational complexity and redundancy) or too few (insufficient accuracy and unsolvable solutions), and avoiding over-reliance on a single satellite system (such as selecting only GPS satellites). This enhances the system's "interference immunity" and "compatibility."

[0056] function Relying on real-time satellite status, Constraints can be set as needed. In occluded scenes, the quantity penalty weight is appropriately reduced to prioritize solution continuity; in open scenes, the distribution penalty weight is increased to focus on high-precision combinations, allowing the evaluation logic to adapt to dynamic environments.

[0057] As a preferred embodiment:

[0058] In the above formula, Represents the trace operation of the matrix, which is used to convert the result of the matrix inverse into a scalar variance; Represents the design matrix Perform transpose operation; represents the weight matrix; Design Matrix

[0059] In the above formula, represents the gradient of the double-difference carrier phase observation; Represents the phase change rate versus heading angle The partial derivative of represents the wavelength-dependent coefficient, which is used to convert phase changes into spatial geometric quantities; , Indicates the The horizontal component of the double difference pair is used to describe the geometric distribution of satellites on the horizontal plane. 、 The m in represents the number of double-difference observations formed by the selected satellite combination S; Weight Matrix ; In the above formula, Represents the construction of a diagonal matrix; m represents the number of double-difference observations composed of the selected satellite combination S; No. double-difference observations The weight of is calculated as follows: ; In the above formula, Indicates the The satellite number corresponding to the double difference pair; 、 Respectively represent The carrier-to-noise ratio of the two satellites in a double-difference pair; 、 Respectively represent The phase observation noise variance of the two satellites in a double-difference pair.

[0060] Design Matrix It is the bridge between “satellite geometric distribution” and “phase observation”. The more dispersed the geometric distribution ( , The difference is large), The higher the rank, the greater the potential accuracy of the heading angle solution.

[0061] As a preferred embodiment: The constraint function for reducing the ambiguity search space based on the baseline length constraint is as follows: ; In the above formula, represents the double-difference integer ambiguity; Indicates the baseline length of the dual antenna; Indicates the satellite carrier wavelength; Indicates the maximum possible angle difference between satellites; As a beneficial effect of the above embodiment: using the baseline length , carrier wavelength The maximum angle difference between the satellite , construct double-difference integer ambiguity Constraint relationship Compressing the ambiguity search space from a theoretically large range to a smaller range related to the actual antenna baseline and satellite distribution significantly reduces the number of candidates for ambiguity resolution and improves resolution efficiency. This is especially true in complex scenarios (such as urban canyons with variable satellite angles), making ambiguity resolution more focused and efficient. Even in environments with obstructed satellite signals and high observation noise, the prior information on baseline length can be used to provide stable constraints for ambiguity fixation, improving the success rate of the subsequent LAMBDA algorithm and ensuring the continuity of heading angle resolution.

[0062] The constraint function for narrowing the ambiguity search space based on the condition number constraint is as follows: ; In the above formula, Represents the condition number of the matrix, which is used to measure the degree of "ill-conditioning" of the matrix; , the specific content is the same as above; Represents the weight matrix, the specific content is the same as above; Represents the upper limit of the matrix condition number, which is a pre-set threshold.

[0063] As a beneficial effect of the above embodiment: the condition number Measure the degree of "ill-conditioning" of the matrix by constraining , can avoid the serious ill-conditioning (excessive condition number) of the matrix due to poor satellite geometric distribution (such as satellite concentration) and large heterogeneity of observation noise. After the matrix ill-conditioning is reduced, the heading angle variance The calculation is more stable and the solution results are more reliable, reducing the heading accuracy jump caused by unstable matrix inversion.

[0064] As a preferred embodiment of step S5, the enhanced LAMBDA algorithm includes the following contents: Modified integer least squares criterion: To make The smallest integer vector, ; In the above formula, represents the final selected integer ambiguity solution; represents an integer ambiguity candidate; represents the floating-point ambiguity solution; represents the covariance matrix; The validation statistics are as follows: ; In the above formula, represents the optimal integer solution; represents a suboptimal integer solution; represents the weighted norm; By calculating the ratio of the weighted distance between the optimal solution and the floating-point solution to the weighted distance between the suboptimal solution and the floating-point solution, the larger the ratio, the better the optimal solution is and the more reliable the integer solution is.

[0065] The conditions for acceptance are as follows: ; In the above formula, Represents the preset threshold, which is the ambiguity resolution reliability verification threshold.

[0066] Only when the "advantage of the optimal solution is large enough" ( ), only then will the integer solution be accepted as the correct solution. Otherwise, the ambiguity fixation is considered unreliable and needs to be recalculated or the observation adjusted.

[0067] The traditional LAMBDA algorithm uses only integer least squares filtering. When floating-point solutions are affected by noise, incorrect integer solutions may be selected (especially in scenarios with poor satellite geometry and high observation noise). This algorithm, by designing an enhanced LAMBDA algorithm, adds a "reliability check" to integer solutions by verifying statistics and acceptance conditions, thus preventing incorrect fixes caused by noise.

[0068] This can improve the robustness in complex scenarios, such as urban canyons, tree-lined avenues and other occluded scenes, where the satellite geometry is poor ( The covariance matrix is ​​ill-conditioned), floating-point solutions have high uncertainty, and observation noise (multipath, weak signals) can easily lead the least squares criterion to mistakenly select a suboptimal solution. The enhanced LAMBDA algorithm of this invention can identify such situations where the optimal solution is not advantageous enough, reject unreliable integer solutions, avoid large deviations in heading angle calculations due to ambiguity errors, and improve the robustness of the system in complex scenarios.

[0069] As a preferred embodiment of step S6, the algorithm for updating the trigger mechanism is as follows: Construct an environmental change indicator function:

[0070] Update condition determination: if or is an integer, then,

[0071] In other cases, ; In the above formula, is the change threshold; For regular update cycles; An indicator that represents the relative magnitude of environmental change; represents the environmental state vector of the dual-antenna direction finding system at time t in step S1; Represents the environmental state vector before T time; When the relative magnitude of environmental change exceeds the threshold ( ), it is determined that "the environment has changed significantly" and the parameter update is triggered (execute steps S1 to S4 and re-screen satellites). A typical scenario is: for example, a vehicle drives from an open area into an urban canyon.

[0072] Even if the environment appears stable ( ), and also at a fixed period Force parameter updates to cope with "slow environmental changes" (such as gradually entering mountainous areas and the number of satellites slowly decreasing) to avoid long-term non-updates of parameters due to changes not reaching the threshold.

[0073] Through the above settings, real-time monitoring of environmental change indicators , ensuring that parameters can be adjusted quickly in scenarios such as tunnels and high-rise building obstructions, avoiding system failure due to "using old parameters to cope with new environments" and improving adaptability in complex environments.

[0074] After the update trigger mechanism is started in step S6, the algorithm uses a smooth update algorithm to prevent the altitude angle threshold, the carrier-to-noise ratio threshold and the minimum number of satellites from jumping directly to the target value.

[0075] The smooth update algorithm is as follows: ; In the above formula, represents the update step size; Indicates the current moment ( step) parameter vector (such as altitude angle threshold, carrier-to-noise ratio threshold, minimum number of satellites, etc.); Indicates the next moment ( step) parameter vector; represents the target parameter vector; The component constraint algorithm is as follows:

[0076] In the above formula, =1,2,3, Indicates the maximum change in the altitude angle threshold, Indicates the maximum change in the carrier-to-noise ratio threshold. Indicates the maximum change in the minimum number of satellites.

[0077] By setting a limit on the "single change amplitude" of each parameter component, we can avoid a parameter (such as the altitude angle threshold) from "instantaneously jumping" due to sudden environmental changes, which would cause system output oscillation (such as a sudden and large change in the heading angle).

[0078] When the environment suddenly changes (e.g., entering a tunnel, or the number of satellites suddenly drops), target parameters may change significantly. Smooth updates combined with component constraints allow for "slow adjustment" of parameters, preventing wild fluctuations in heading angle and ambiguity resolution due to parameter jumps (e.g., a sudden 10° jump in vehicle heading). Parameters gradually approach target values, responsive to environmental changes (e.g., gradually lowering the elevation angle threshold to include more low-elevation-angle satellites) while maintaining stable system output. This approach is particularly suitable for dynamic vehicles, such as vehicles and aircraft, that require high-precision navigation.

[0079] After the selected satellite set S is determined in step S4, the heading angle is calculated and output using the following algorithm. Specifically: The heading angle is estimated by the least squares method: ; In the above formula, represents the estimated heading angle; Indicates satellite The weight of 、 Indicates satellite horizontal component of the double difference pair; Indicates satellite Double-difference carrier phase observation value; For heading accuracy estimation, the heading angle standard deviation function is as follows: ; In the above formula, represents the standard deviation of the heading angle; Indicates satellite The weight of Indicates satellite The horizontal component x of the double difference pair; By summing all satellites in the selected satellite set S, the standard deviation of the heading angle is accurately calculated; Confidence interval: ; where t represents the critical value of the distribution (given by the confidence level and degrees of freedom Sure) Confidence interval = heading angle mean ± probability coefficient × standard deviation. Using the standard deviation and confidence interval, the heading angle solution is upgraded from a "single value" to a complete output of "result + error probability." This not only answers the question "What is the heading angle?" but also the question "How accurate is it?" This system is key to the engineering application of dual-antenna direction-finding systems and is suitable for complex scenarios (such as high-speed vehicle movement) where rapid calculation of standard deviations and confidence intervals is required. This can be achieved through fast calculation using pre-stored satellite distribution templates and weights.

[0080] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise", "axial", "radial", "circumferential" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0081] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, features specified as "first" or "second" may explicitly or implicitly include one or more of the features.

[0082] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A GNSS dual-antenna direction finding adaptive satellite screening algorithm, characterized in that: The following steps are involved: S1. Initialize the altitude angle threshold and the carrier-to-noise ratio threshold, and based on the initial values ​​of the altitude angle threshold and the carrier-to-noise ratio threshold, screen out all satellites that meet the visibility conditions; S2. Based on all satellites that meet the visibility conditions, construct an environmental state vector of the dual-antenna direction finding system at time t. The environmental state vector includes the number of satellites that meet the visibility conditions, the average carrier-to-noise ratio of the satellites, the standard deviation of the satellite carrier-to-noise ratio, and the average satellite elevation angle; S3. Build an adaptive function that dynamically adjusts the altitude angle threshold and the carrier-to-noise ratio threshold according to the environmental state vector, and re-screens candidate satellites that meet the threshold requirements based on the adjusted altitude angle threshold and the carrier-to-noise ratio threshold; S4. Construct a multi-dimensional satellite comprehensive scoring function, evaluate the quality of all candidate satellites in step S3 using the comprehensive scoring function, and eliminate satellites with too low scores; S5. The remaining satellites in step S4 are screened out to obtain the optimal satellite set through the heading accuracy evaluation function.

2. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 1, characterized in that: The step S6 is also included: S6. Perform double-difference ambiguity resolution optimization on the satellite set selected in step S5, narrow the ambiguity search space based on baseline length and condition number constraints, and use the enhanced LAMBDA algorithm to solve it. Calculate the heading angle based on the solution result and output it.

3. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 1, characterized in that: The step S7 is also included: S7. Establish a dynamic parameter update mechanism to re-evaluate the environmental status and adjust relevant parameters when significant changes in the environment are detected.

4. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 1, characterized in that: In step S1, the visibility determination function , used to determine whether a single satellite at time t meets the visibility condition; Visibility determination function as follows: if ,but ,but ; In the above formula, Indicates that at time t the satellite Altitude angle; represents the altitude mask, Indicates that at time t the satellite carrier-to-noise ratio; represents the carrier-to-noise ratio mask; In step S2, the environment state vector is as follows: ； represents the environmental state vector of the dual-antenna direction finding system at time t, where the parameters in the environmental state vector are calculated using the following formula: , represents the number of satellites that meet the visibility conditions at time t; , It represents the average carrier-to-noise ratio of all satellites that meet the visibility conditions at time t, which is used to reflect the overall strength level of the current satellite signal; , It represents the standard deviation of the carrier-to-noise ratio of all satellites that meet the visibility conditions at time t, and is used to reflect the dispersion of the carrier-to-noise ratio of each satellite; , It represents the average elevation angle of all satellites that meet the visibility conditions at time t, and is used to reflect the overall elevation angle distribution of satellites in the sky.

5. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 4, characterized in that: In step S3, the adaptive function includes an altitude angle threshold adaptive function, and the altitude angle threshold adaptive function is as follows: ； In the above formula, Indicates the final altitude angle threshold for satellite screening, its value depends on the number of currently visible satellites ; Indicates the initial value of the altitude angle threshold; Indicates the adaptive adjustment amount of the altitude angle threshold.

6. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 4, characterized in that: In step S3, the adaptive function includes a carrier-to-noise ratio threshold adaptive function, and the carrier-to-noise ratio threshold adaptive function is as follows: ； In the above formula, Indicates the carrier-to-noise ratio threshold ultimately used for satellite screening; Indicates the initial value of the carrier-to-noise ratio threshold; The weight of the impact of the number of satellites on the adjustment of the carrier-to-noise ratio threshold; represents the satellite number adjustment function; Indicates the impact weight of environmental quality; represents the environmental quality adjustment function; ； In the above formula, represents the hyperbolic tangent function; Indicates the number of satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the number of satellite reference points; A indicates the gain factor; B indicates the scaling factor; C indicates the offset constant; In the above formula, D represents the gain factor; Indicates the environmental quality benchmark point; In the above formula, 、 Respectively represent the weights of environmental quality dimensions; Represents the average carrier-to-noise ratio of all satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the normalized maximum value of the carrier-to-noise ratio; It represents the standard deviation of the carrier-to-noise ratio of all satellites that meet the visibility conditions, obtained through the environmental state vector; Indicates the normalized maximum value of the standard deviation of the carrier-to-noise ratio; Represents the average elevation angle of all satellites that meet the visibility conditions, obtained through the environmental state vector.

7. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 1, characterized in that: In step S4, the comprehensive scoring function of a single satellite is as follows: ； In the above formula, It indicates that each satellite is given a comprehensive score by weighted summing up the scores of six dimensions: carrier-to-noise ratio, elevation angle, azimuth diversity, carrier phase quality, multipath effect, and double-difference residual; Represents the weight vector of each dimension; Indicates the Satellite No. Ratings of dimensions.

8. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 7, characterized in that: The sub-functions in the comprehensive scoring function are as follows: The carrier-to-noise ratio scoring function is: ,in, Indicates that at time t The carrier-to-noise ratio of the satellite; The altitude angle scoring function is: ,in, Indicates that at time t The altitude angle of the satellite; The azimuthal diversity scoring function is: ,in, Indicates that at time t The azimuth of the satellite; , represents the azimuth of all satellites that meet the visibility conditions at time t, It means that at time t, all satellites that meet the visibility conditions are summed up to obtain the total azimuth angle. represents all satellites that meet the visibility conditions at time t; The carrier phase quality scoring function is: ,in, Indicates that at time t The phase observation noise of the satellites, represents the standard deviation of the double-difference carrier phase noise at time t; The multipath effect scoring function is: ,in, represents the multipath interference intensity of the i-th satellite at time t; , represents the pseudorange observation value of the i-th satellite at time t, represents the geometric distance of the i-th satellite at time t, represents the speed of light; The double-difference residual score function is: ,in, represents the absolute value of the double-difference observation residual of the i-th satellite at time t, represents the a priori standard deviation of the double-differenced residuals.

9. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 1, characterized in that: In step S5, the heading accuracy evaluation function is as follows: ； In the above formula, represents the selected satellite set; represents the heading angle variance; represents the weight coefficient; Represents a penalty term.

10. The GNSS dual-antenna direction finding adaptive satellite screening algorithm according to claim 2, characterized in that: The constraint function for reducing the ambiguity search space based on the baseline length constraint is as follows: ； In the above formula, represents the double-difference integer ambiguity; Indicates the baseline length of the dual antenna; Indicates the satellite carrier wavelength; Indicates the maximum possible angle difference between satellites; The constraint function for narrowing the ambiguity search space based on the condition number constraint is as follows: ； In the above formula, represents the condition number of the matrix; Represents the design matrix Perform transpose operation; represents the weight matrix; represents the upper limit of the matrix condition number.

Images