Embedded AGV positioning and navigation method based on differential GNSS

Abstract

The application relates to the technical field of positioning and navigation, and discloses an embedded AGV positioning and navigation method based on differential GNSS, which comprises the following steps: arranging a plurality of passive GNSS patch antennas on the top of a vehicle in a circumferential manner; combining the carrier-to-noise ratio and the elevation angle information output by double receivers to construct a low-elevation satellite set; calculating the carrier-to-noise ratio difference on the symmetrical antennas; and performing periodic analysis to extract the environmental shielding law and divide phase intervals. The system can dynamically select the optimal antenna combination in each phase interval, and realizes real-time differential navigation solution in the navigation operation, so that the signal attenuation and navigation interruption problems caused by the narrow lane shielding can be effectively overcome, and the high-precision, stable and continuous navigation of the AGV in a complex environment can be ensured.

Description

Technical Field

[0001] This invention relates to the field of positioning and navigation technology, and more specifically, to an embedded AGV positioning and navigation method based on differential GNSS. Background Technology

[0002] Differential GNSS technology can provide correction information through a base station, enabling mobile devices to obtain high-precision positioning results in complex environments. For industrial automated vehicles, the commonly used dual-antenna moving baseline RTK method can not only acquire position but also calculate heading in real time, thus achieving attitude awareness without the need for a magnetic compass. This method has the advantages of high accuracy, fast response, and ease of embedded implementation in open environments, and is therefore widely used in the positioning and navigation of automated guided vehicles (AGVs).

[0003] However, in the narrow aisle environment of warehousing and logistics, metal shelves are densely arranged with regularly spaced uprights. In this scenario, GNSS signals are highly susceptible to obstruction and reflection, especially satellite signals from low elevation directions, which are more easily lost. Since AGVs typically travel along the aisle at a constant speed, the shelf uprights create a regular obstruction effect, causing periodic signal attenuation or loss on one side of the antenna. As a result, the number of observations of the common satellite by both antennas exhibits a regular fluctuation of decrease and recovery over time, while heading calculation must rely on a sufficient number of common-view satellites.

[0004] Existing dual-antenna differential GNSS heading calculation methods do not consider the periodic occlusion effect in the warehouse environment. Once the number of common-view satellites is insufficient, the heading calculation will be frequently interrupted or abruptly changed, causing discontinuous heading output of AGV during navigation, trajectory jitter, and even affecting the vehicle's path following and safety. Summary of the Invention

[0005] This invention provides an embedded AGV positioning and navigation method based on differential GNSS, which solves the technical problems mentioned in the background art.

[0006] This invention provides an embedded AGV positioning and navigation method based on differential GNSS, comprising:

[0007] Multiple passive global satellite navigation system patch antennas are evenly distributed along the circumference of the vehicle's roof;

[0008] The radio frequency inputs of the two global satellite navigation system receivers are connected to any patch antenna, and the outputs are the carrier-to-noise ratio and elevation angle of each satellite;

[0009] Determine the set of low-elevation satellites based on the elevation angles of all connected satellites;

[0010] Calculate the carrier-to-noise ratio difference for a low-elevation-angle satellite array on two patch antennas symmetrically arranged on a circle.

[0011] The environmental occlusion period is obtained by performing periodic analysis on the carrier-to-noise ratio difference, and the phase interval is divided.

[0012] Within each phase interval, the first and second antennas corresponding to the low-elevation-angle satellite set are statistically analyzed.

[0013] During the positioning and navigation operation period, the corresponding phase interval is determined to obtain the corresponding first antenna and second antenna;

[0014] Real-time dynamic differential heading calculation is performed based on the first and second antennas.

[0015] Furthermore, a set of low-elevation satellites is determined based on the elevation angles of all connected satellites, including:

[0016] At time ti, obtain the first set of satellites and the second set of satellites whose elevation angles are simultaneously output by the receivers of the two global satellite navigation systems. Calculate the intersection of the first set of satellites and the second set of satellites to obtain the common connected satellite set.

[0017] Extract the elevation angles of all satellites in the common connected satellite set and sort them in ascending order to obtain the elevation angle sequence;

[0018] The average elevation angle of the elevation angle sequence is used to obtain the segmentation threshold;

[0019] If the elevation angle of the i-th satellite in the common connected satellite set is less than or equal to the segmentation threshold, it is marked as a low elevation angle satellite to form a low elevation angle satellite set.

[0020] Furthermore, on two symmetrical patch antennas on the circumference, the carrier-to-noise ratio difference of the low-elevation satellite array is calculated, including:

[0021] Determine any pair of symmetrical patch antennas to form an antenna pair;

[0022] The carrier-to-noise ratio of each satellite in the low-elevation satellite array is synchronously acquired by RF multiplexing.

[0023] Calculate the average carrier-to-noise ratio (CNR) of each patch antenna pair for each satellite in the antenna pair to obtain the first average CNR and the second average CNR.

[0024] The difference between the first average carrier-to-noise ratio and the second average carrier-to-noise ratio is calculated to obtain the corresponding carrier-to-noise ratio difference.

[0025] Furthermore, periodic analysis of the carrier-to-noise ratio difference yields the environmental occlusion period, including:

[0026] Based on the carrier-to-noise ratio (CNR) difference of the antenna pair at N time points, a CNR difference sequence is formed;

[0027] Calculate the mean of the carrier-to-noise ratio difference sequence to remove the mean from the carrier-to-noise ratio difference sequence, and obtain the mean-removed carrier-to-noise ratio difference sequence.

[0028] If the i-th de-average load ratio difference in the de-average load ratio difference sequence is less than the (i-1)-th de-average load ratio difference, and the i-th de-average load ratio difference is less than the (i+1)-th de-average load ratio difference, then the i-th de-average load ratio difference is marked as a local minimum to form a set of local minima.

[0029] If the i-th de-average load noise ratio difference in the de-average load noise ratio difference sequence is greater than the (i-1)-th de-average load noise ratio difference, and the i-th de-average load noise ratio difference is greater than the (i+1)-th de-average load noise ratio difference, then the i-th de-average load noise ratio difference is marked as a local maximum to form a set of local maxima.

[0030] Calculate the autocorrelation function of the de-average noise ratio difference sequence, select the de-average noise ratio difference corresponding to the first local maximum, and obtain the environmental occlusion period.

[0031] Furthermore, the phase intervals are divided, including:

[0032] Determine the position of each de-average load noise ratio difference in the local minimum set in the de-average load noise ratio difference sequence, and select the time corresponding to the de-average load noise ratio difference with the smallest position as the reference time.

[0033] Determine the time corresponding to each difference in the average load-to-noise ratio in the local minimum set as the boundary time;

[0034] Calculate the ratio of the difference between the boundary time and the reference time to the environmental occlusion period, and map the ratio to obtain the corresponding normalized time phase through the frac function;

[0035] Arrange each normalized time phase in ascending order to obtain the normalized time phase sequence;

[0036] Calculate the average value of adjacent normalized time phases in the normalized time phase sequence, and sort the average values ​​in ascending order to obtain the average value sequence;

[0037] Extract the phase intervals corresponding to adjacent average values ​​in the average value sequence.

[0038] Furthermore, within each phase interval, the first and second antennas corresponding to the low-elevation satellite set are statistically analyzed, including:

[0039] For each phase interval in the phase interval set, determine each moment when the normalized time phase belongs to the phase interval, forming a time set;

[0040] For each time u in the time set and each patch antenna k: the sum of the carrier-to-noise ratios of each satellite in the low-elevation satellite set at time u on the patch antenna k is divided by the number of satellites in the low-elevation satellite set to obtain the average carrier-to-noise ratio of the patch antenna k at the low-elevation angle at time u.

[0041] The sum of the average carrier-to-noise ratio at low elevation angle for each time in the time set is divided by the number of time points in the time set to obtain the statistics of patch antenna k in the phase interval.

[0042] Select the patch antenna with the largest statistic from all patch antennas as the first patch antenna corresponding to the phase interval;

[0043] The patch antenna symmetrical to the first patch antenna is used as the second patch antenna.

[0044] Furthermore, real-time dynamic differential heading calculation is performed based on the first and second antennas, including:

[0045] Connect the first patch antenna to receiver A and the second patch antenna to receiver B;

[0046] At the current moment, the same differential correction data is simultaneously injected into receiver A and receiver B, and receiver A and receiver B are forced to perform the calculation with the current moment as the time stamp;

[0047] Receiver A outputs the current east and north coordinates, and receiver B outputs the current east and north coordinates.

[0048] The eastward coordinate difference is calculated as the eastward coordinate of receiver B minus the eastward coordinate of receiver A, and the northward coordinate difference is calculated as the northward coordinate of receiver B minus the northward coordinate of receiver A.

[0049] The baseline length is the square root of the sum of the squares of the eastward coordinate difference and the squares of the northward coordinate difference;

[0050] The heading angle is the target angle calculated by using the two-parameter arctangent function on the eastward coordinate difference and the northward coordinate difference. The target angle is the angle between the vector formed by the northward clockwise direction and the vector composed of the eastward coordinate difference and the northward coordinate difference.

[0051] The beneficial effects of this invention are as follows: Addressing the periodic obstruction problem caused by metal shelves in narrow aisle environments of warehousing and logistics, this invention effectively extracts and utilizes low-elevation-angle satellite signal characteristics through multi-antenna deployment, carrier-to-noise ratio difference analysis, and phase interval division. This allows for dynamic selection of the optimal antenna combination and differential heading calculation under complex obstruction conditions. This not only avoids the heading calculation interruptions and trajectory jitter caused by insufficient common-view satellites in traditional dual-antenna methods, but also significantly improves the continuity and stability of AGV heading output, thereby ensuring high-precision, safe, and reliable path following and automatic navigation for vehicles in complex environments. Attached Figure Description

[0052] Figure 1 This is a flowchart of the embedded AGV positioning and navigation method based on differential GNSS of the present invention. Detailed Implementation

[0053] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0054] like Figure 1 As shown, the embedded AGV positioning and navigation method based on differential GNSS includes:

[0055] Multiple passive global satellite navigation system patch antennas are evenly distributed along the circumference of the vehicle's roof;

[0056] The radio frequency inputs of the two global satellite navigation system receivers are connected to any patch antenna, and the outputs are the carrier-to-noise ratio and elevation angle of each satellite;

[0057] Determine the set of low-elevation satellites based on the elevation angles of all connected satellites;

[0058] Calculate the carrier-to-noise ratio difference for a low-elevation-angle satellite array on two patch antennas symmetrically arranged on a circle.

[0059] The environmental occlusion period is obtained by performing periodic analysis on the carrier-to-noise ratio difference, and the phase interval is divided.

[0060] Within each phase interval, the first and second antennas corresponding to the low-elevation-angle satellite set are statistically analyzed.

[0061] During the positioning and navigation operation period, the corresponding phase interval is determined to obtain the corresponding first antenna and second antenna;

[0062] Real-time dynamic differential heading calculation is performed based on the first and second antennas.

[0063] In one embodiment of the present invention, determining the low-elevation satellite set based on the elevation angles of all connected satellites includes:

[0064] At time ti, obtain the first set of satellites and the second set of satellites whose elevation angles are simultaneously output by the receivers of the two global satellite navigation systems. Calculate the intersection of the first set of satellites and the second set of satellites to obtain the common connected satellite set.

[0065] Extract the elevation angles of all satellites in the common connected satellite set and sort them in ascending order to obtain the elevation angle sequence;

[0066] The average elevation angle of the elevation angle sequence is used to obtain the segmentation threshold;

[0067] If the elevation angle of the i-th satellite in the common connected satellite set is less than or equal to the segmentation threshold, it is marked as a low elevation angle satellite to form a low elevation angle satellite set.

[0068] In detail, at time ti, the two sets of global navigation satellite system receivers output their respective sets of observed satellites, namely the first set of satellites and the second set of satellites. The first set of satellites consists of satellites whose signals can be stably received by the first set of receivers at that time, and the second set of satellites consists of satellites whose signals can be stably received by the second set of receivers at the same time. Since the dual-antenna azimuth calculation depends on the joint observation of the same satellite by both antennas, the signal can only be used for subsequent differential analysis when the satellite is simultaneously acquired by both sets of receivers. Therefore, it is necessary to calculate the intersection of the two sets to obtain the common connected satellite set. The common connected satellite set includes satellites that can be observed by both sets of receivers at time ti.

[0069] In detail, the elevation angle data of each satellite is extracted from the set of publicly connected satellites. The elevation angle refers to the angle between the line connecting the satellite and the observation point (i.e., the receiver) and the ground plane, reflecting the satellite's altitude in the sky. The elevation angles are arranged in ascending order to form an elevation angle sequence. This ordered processing facilitates statistical analysis of the distribution characteristics of satellites at different altitudes.

[0070] In detail, based on the sorted elevation angle sequence, the average elevation angle of all elevation angles is calculated, and this average elevation angle is used as the segmentation threshold. The average elevation angle is chosen as the threshold because it can adapt to the satellite elevation angle distribution characteristics under different environments. In a narrow alleyway environment, affected by the obstruction of metal shelves on both sides, the overall elevation angle distribution of visible satellites may differ from that in an open environment. A fixed threshold is difficult to adapt to dynamically changing scenarios, while the average elevation angle can dynamically reflect the overall altitude level of currently connected satellites, making the definition of low elevation angles more consistent with the real-time environment.

[0071] In detail, for each satellite in the common connected satellite set, its elevation angle is compared with a segmentation threshold. If a satellite's elevation angle is less than or equal to the segmentation threshold, it is marked as a low-elevation satellite. All marked low-elevation satellites together constitute the low-elevation satellite set.

[0072] Low-elevation satellites (at a lower altitude) are closer to the metal shelves on both sides in narrow aisles, and their signals are more easily blocked by the shelves or affected by multipath reflections. The fluctuation of their C / N0 value can more sensitively reflect the periodic blocking characteristics in the aisle (such as the blocking changes caused by the shelf pitch).

[0073] In one embodiment of the present invention, calculating the carrier-to-noise ratio difference of a low-elevation-angle satellite array on two circumferentially symmetrical patch antennas includes:

[0074] Determine any pair of symmetrical patch antennas to form an antenna pair;

[0075] The carrier-to-noise ratio of each satellite in the low-elevation satellite array is synchronously acquired by RF multiplexing.

[0076] Calculate the average carrier-to-noise ratio (CNR) of each patch antenna pair for each satellite in the antenna pair to obtain the first average CNR and the second average CNR.

[0077] The difference between the first average carrier-to-noise ratio and the second average carrier-to-noise ratio is calculated to obtain the corresponding carrier-to-noise ratio difference.

[0078] In detail, among multiple patch antennas arranged in a circle, any pair of two patches that are geometrically symmetrically distributed are selected to form an antenna pair. Since the sides of the aisle are usually lined with regularly arranged metal shelves, the symmetrical antenna pairs can correspond to the directions on both sides of the aisle, and the difference in their signal quality can directly reflect the different degrees of obstruction on both sides. For example, if a pair of symmetrical antennas are directed towards the shelves on the left and right sides of the aisle respectively, the difference in their carrier-to-noise ratio can sensitively indicate which side of the shelf is more significantly obstructed.

[0079] In detail, the RF multiplexer is used to ensure that the carrier-to-noise ratio (CNR) data for each satellite in the low-elevation satellite array is acquired simultaneously by the two patch antennas. CNR is a key indicator of satellite signal quality; a lower value indicates more severe signal interference from blockages, multipath propagation, or noise. In narrow aisle environments, CNR fluctuates dynamically with the relative position of the antenna and the rack; synchronous acquisition eliminates errors caused by time differences.

[0080] In detail, for the first patch in the antenna pair, the average carrier-to-noise ratio (CNR) of the patch with respect to all satellites in the low-elevation satellite array is calculated to obtain the first average CNR. Similarly, the average CNR of the second patch with respect to all satellites in the array is calculated to obtain the second average CNR. Using the average value instead of the CNR of a single satellite is to suppress random noise interference from individual satellite signals. A single satellite in the low-elevation satellite array may experience CNR jumps due to momentary obstruction or multipath propagation, while the average value smooths out these random fluctuations, more stably reflecting the overall reception quality of the antenna in the low-elevation region.

[0081] In detail, the carrier-to-noise ratio (CNR) difference is obtained by subtracting the first average CNR from the second average CNR. This CNR difference is used to quantify the signal quality differences of symmetrical antennas in low-elevation-angle sensitive areas: a positive difference indicates that the average signal quality of the first patch is better than the second, possibly due to less obstruction on the side it faces; a negative difference indicates that the second patch is more severely affected by obstruction or multipath propagation. In narrow aisles, the CNR difference changes periodically with the rack pitch as the equipment moves. When an antenna on one side approaches a rack column, its average CNR decreases, and the CNR difference fluctuates regularly.

[0082] In one embodiment of the present invention, the environmental occlusion period is obtained by performing periodic analysis on the carrier-to-noise ratio difference, including:

[0083] Based on the carrier-to-noise ratio (CNR) difference of the antenna pair at N time points, a CNR difference sequence is formed;

[0084] Calculate the mean of the carrier-to-noise ratio difference sequence to remove the mean from the carrier-to-noise ratio difference sequence, and obtain the mean-removed carrier-to-noise ratio difference sequence.

[0085] If the i-th de-average load ratio difference in the de-average load ratio difference sequence is less than the (i-1)-th de-average load ratio difference, and the i-th de-average load ratio difference is less than the (i+1)-th de-average load ratio difference, then the i-th de-average load ratio difference is marked as a local minimum to form a set of local minima.

[0086] If the i-th de-average load noise ratio difference in the de-average load noise ratio difference sequence is greater than the (i-1)-th de-average load noise ratio difference, and the i-th de-average load noise ratio difference is greater than the (i+1)-th de-average load noise ratio difference, then the i-th de-average load noise ratio difference is marked as a local maximum to form a set of local maxima.

[0087] Calculate the autocorrelation function of the de-average noise ratio difference sequence, select the de-average noise ratio difference corresponding to the first local maximum, and obtain the environmental occlusion period.

[0088] In detail, the carrier-to-noise ratio (CNR) difference of the antenna pair at N consecutive moments is arranged in chronological order to form a CNR difference sequence. In a narrow aisle environment, when the AGV vehicle travels at a constant speed along the aisle, the pillars or grids of the shelves on both sides will periodically block the antenna's field of view at fixed intervals, causing the CNR difference of the symmetrical antenna pair to fluctuate regularly over time.

[0089] In detail, the mean of all elements in the carrier-to-noise ratio (CNR) difference sequence is calculated, and then this mean is subtracted from the CNR difference at each time step in the sequence to obtain the mean-reduced CNR difference sequence. This eliminates the DC component (i.e., the overall offset) in the sequence, highlighting the fluctuations. In real-world scenarios, the CNR difference may have a constant deviation due to differences in antenna hardware or overall environmental offsets. After removing the mean, the fluctuations in the sequence will more purely reflect the dynamic changes caused by periodic obstruction, avoiding the constant deviation from masking the periodic characteristics.

[0090] In detail, for the sequence of carrier-to-noise ratio (CNR) differences, local extrema are determined point by point: if the i-th element is less than its preceding element (i-1) and less than its following element (i+1), then this element is a local minimum and is included in the set of local minima; if the i-th element is greater than its preceding element and greater than its following element, then this element is a local maximum and is included in the set of local maxima. Local extrema are key characteristic points of periodic fluctuations. In a narrow aisle, when equipment travels to a position where a shelf column faces an antenna on one side, the CNR of that antenna drops sharply, and the CNR difference will exhibit a local minimum (if that side is the first antenna) or a maximum (if that side is the second antenna); while when the equipment is located in the gap between the columns, the CNR difference will fluctuate in the opposite direction, forming an extremum.

[0091] In detail, the autocorrelation function is used to measure the similarity of sequences at different time lags. For periodic sequences, the autocorrelation function peaks at lags that are integer multiples of the period. The autocorrelation function of the de-average load-to-noise ratio difference sequence is calculated; each value of this function corresponds to the similarity between the sequence and its offset at a certain lag. The lag corresponding to the first local maximum in the autocorrelation function is selected; this lag represents the environmental occlusion period. The reason for choosing the first local maximum is that it corresponds to the basic period of the sequence (i.e., the minimum repetition period determined by the shelf pitch and equipment speed), while higher-order maxima may correspond to multiples of the period, not the actual physical period. This allows for the extraction of the time scale of periodic occlusion in narrow aisle environments based on time-series data.

[0092] In one embodiment of the present invention, dividing the phase interval includes:

[0093] Determine the position of each de-average load noise ratio difference in the local minimum set in the de-average load noise ratio difference sequence, and select the time corresponding to the de-average load noise ratio difference with the smallest position as the reference time.

[0094] Determine the time corresponding to each difference in the average load-to-noise ratio in the local minimum set as the boundary time;

[0095] Calculate the ratio of the difference between the boundary time and the reference time to the environmental occlusion period, and map the ratio to obtain the corresponding normalized time phase through the frac function;

[0096] Arrange each normalized time phase in ascending order to obtain the normalized time phase sequence;

[0097] Calculate the average value of adjacent normalized time phases in the normalized time phase sequence, and sort the average values ​​in ascending order to obtain the average value sequence;

[0098] Extract the phase intervals corresponding to adjacent average values ​​in the average value sequence.

[0099] In detail, from the set of local minima, the position (i.e., chronological order) of each local minimum difference in the sequence of local minimum differences is extracted, and the time corresponding to the local minimum with the smallest position is selected as the reference time. The valley of the local minimum difference corresponds to the lowest point of the carrier-to-noise ratio difference, reflecting the moment when the antenna on a certain side is most severely blocked by the rack, which is a key characteristic point of periodic blockage. Selecting the earliest local minimum time as the reference establishes a unified phase starting point, ensuring the consistency of subsequent phase calculations in the time dimension and avoiding phase confusion caused by an unstable starting point.

[0100] In detail, the moments corresponding to all differences in the average load-to-noise ratio within the local minima are taken as boundary moments. In a narrow aisle environment, the interval between adjacent local minima corresponds to a complete occlusion cycle (e.g., the time from one occlusion to the next for a shelf upright), and these moments are nodes where the occlusion state repeatedly changes. Using these nodes as boundaries, the continuous time axis can be divided into segments synchronized with the occlusion cycle, with each segment containing a complete "occlusion → mitigation → re-occlusion" process.

[0101] In detail, for each boundary moment, the time difference between it and the reference moment is calculated. This difference is then divided by the environmental occlusion period to obtain the ratio of the relative period. The frac function (taking the decimal part of this ratio) is used to map it to the interval [0,1) to obtain the corresponding normalized time phase. This converts absolute time into relative phase within the period, eliminating the influence of device speed fluctuations or small changes in the period. For example, if the environmental occlusion period is T, and the difference between a certain boundary moment and the reference moment is 1.3T, then the normalized phase is 0.3, indicating that this moment is located at 30% of the period, achieving phase uniformity for the same occlusion stage within different periods.

[0102] In detail, all normalized time phases are arranged in ascending order of value to form a normalized time phase sequence. This ensures that the phases are arranged in chronological order within the period, reflecting the continuous change of the occlusion state from the reference time. Subsequently, the average value of two adjacent normalized time phases in the sequence is calculated, and these average values ​​are arranged in ascending order to obtain an average value sequence. The average value of adjacent phases is the intermediate phase between two boundary times, corresponding to the intermediate stage of the occlusion state transitioning from one minimum point (the most severe occlusion) to the next minimum point, and can be used as a dividing point for phase intervals.

[0103] In detail, the phase intervals are determined by using two adjacent average values ​​in the average value sequence as endpoints. Each interval corresponds to the phase range between two adjacent boundary moments in the normalized time phase sequence, reflecting a continuous change in occlusion status within the period. For example, if the average value sequence is [0.2, 0.5, 0.8], then the phase intervals are [0, 0.2), [0.2, 0.5), [0.5, 0.8), and [0.8, 1). The occlusion characteristics (such as the degree of occlusion of an antenna on one side) within each interval are similar. The division of these intervals allows the subsequent selection of antenna pairs to match the occlusion status at different stages within the period, ensuring that the antenna pair least affected by occlusion is selected within each phase interval, maximizing the number of satellites viewed by both antennas.

[0104] In one embodiment of the present invention, within each phase interval, the first antenna and the second antenna corresponding to the low-elevation satellite set are statistically analyzed, including:

[0105] For each phase interval in the phase interval set, determine each moment when the normalized time phase belongs to the phase interval, forming a time set;

[0106] For each time u in the time set and each patch antenna k: the sum of the carrier-to-noise ratios of each satellite in the low-elevation satellite set at time u on the patch antenna k is divided by the number of satellites in the low-elevation satellite set to obtain the average carrier-to-noise ratio of the patch antenna k at the low-elevation angle at time u.

[0107] The sum of the average carrier-to-noise ratio at low elevation angle for each time in the time set is divided by the number of time points in the time set to obtain the statistics of patch antenna k in the phase interval.

[0108] Select the patch antenna with the largest statistic from all patch antennas as the first patch antenna corresponding to the phase interval;

[0109] The patch antenna symmetrical to the first patch antenna is used as the second patch antenna.

[0110] In detail, for each phase interval in the phase interval set, all times where the normalized time phase falls within that interval are selected and aggregated into a time set. Each phase interval corresponds to a specific phase within an environmental obstruction period (a transition process between one obstruction peak and the next), and the time set contains all sampling times within that phase. Collecting data from these times allows for centralized analysis of the antenna's performance during the obstruction phase, ensuring that the statistical results accurately reflect the environmental characteristics of that phase interval.

[0111] In detail, for each time *u* in the time set and for all patch antennas *k*, the carrier-to-noise ratio (CNR) of each satellite in the low-elevation satellite set at time *u* on patch antenna *k* is extracted. These CNRs are then summed and divided by the number of satellites in that low-elevation satellite set to obtain the average low-elevation CNR of patch antenna *k* at time *u*. Low-elevation satellites are most sensitive to obstruction and multipath propagation, and their CNR reflects the antenna's reception quality in the current environment. By averaging the CNR of individual satellites, random fluctuations in individual satellite signals can be eliminated, more stably reflecting the antenna's overall reception capability for low-elevation satellites at that time.

[0112] In detail, for each patch antenna k, the average carrier-to-noise ratio at low elevation angles is summed across all times in the time set, and then divided by the number of times in the time set to obtain the statistics for that patch antenna within the current phase interval. These statistics are average performance indicators of the antenna throughout the entire phase interval, smoothing out short-term fluctuations at different times and more reliably reflecting the antenna's stable reception capability during the obstruction phase. A higher statistic indicates less impact from obstruction and multipath propagation on the antenna within that phase interval, resulting in more stable reception quality for low-elevation satellites.

[0113] In detail, from all patch antennas, the patch antenna with the largest statistical value is selected as the first patch antenna corresponding to the current phase interval. The antenna with the largest statistical value is selected because it has the best reception quality for low-elevation satellites in this phase interval, is least affected by obstruction, and can capture low-elevation satellite signals to the maximum extent, providing richer common-view satellite resources for dual-antenna azimuth calculation.

[0114] Specifically, a patch antenna symmetrically distributed circumferentially with the first patch antenna is designated as the second patch antenna. This symmetrical arrangement allows for coverage of both sides of the aisle, ensuring the baseline characteristics of the dual antennas meet the geometric requirements for azimuth calculation. The dual antennas must form an effective baseline to obtain heading information through differential calculation. The symmetrical antenna pair can address the impact of the shelving on both sides at different stages of obstruction. When the first antenna faces the side with less obstruction, the second antenna can cover the other side, thus maintaining a stable number of shared-view satellites.

[0115] In one embodiment of the present invention, real-time dynamic differential heading calculation is performed based on a first antenna and a second antenna, including:

[0116] Connect the first patch antenna to receiver A and the second patch antenna to receiver B;

[0117] At the current moment, the same differential correction data is simultaneously injected into receiver A and receiver B, and receiver A and receiver B are forced to perform the calculation with the current moment as the time stamp;

[0118] Receiver A outputs the current east and north coordinates, and receiver B outputs the current east and north coordinates.

[0119] The eastward coordinate difference is calculated as the eastward coordinate of receiver B minus the eastward coordinate of receiver A, and the northward coordinate difference is calculated as the northward coordinate of receiver B minus the northward coordinate of receiver A.

[0120] The baseline length is the square root of the sum of the squares of the eastward coordinate difference and the squares of the northward coordinate difference;

[0121] The heading angle is the target angle calculated by using the two-parameter arctangent function on the eastward coordinate difference and the northward coordinate difference. The target angle is the angle between the vector formed by the northward clockwise direction and the vector composed of the eastward coordinate difference and the northward coordinate difference.

[0122] In detail, the first patch antenna is connected to receiver A, and the second patch antenna is connected to receiver B. This connection method forms a spatial baseline through two physically separated receiving channels. The geometric distance between the two antennas constitutes the measurement reference, and their relative positional relationship is the basis for heading calculation. In narrow aisle environments, the symmetrically arranged two antennas can sense the signal environment on both sides respectively, and differential processing can cancel common interference, improving the anti-interference capability of heading calculation.

[0123] In detail, at the current moment, the same differential correction data is simultaneously injected into receiver A and receiver B. This differential correction data comes from a reference station and includes correction information for common errors such as satellite orbital error, ionospheric delay, and tropospheric delay. Synchronous injection ensures that both receivers perform calculations based on the same error correction model; it also forces both receivers to perform calculations using the current time as the time stamp, eliminating position deviations caused by time asynchrony. If there is a difference in calculation time, changes in satellite position and signal propagation delay will introduce additional errors; synchronizing the time stamp is crucial to ensuring differential accuracy.

[0124] In detail, receiver A outputs the current eastward and northward coordinates, while receiver B outputs the same eastward and northward coordinates. These coordinates are based on a local coordinate system (usually with east as the x-axis and north as the y-axis) and reflect the positions of the two antennas on the horizontal plane. In real-time dynamic differential mode, the error-corrected coordinate data significantly reduces the impact of common errors and can reflect the relative positional relationship between the two antennas.

[0125] In detail, the eastward coordinate difference is the difference between the eastward coordinates of receiver B and the eastward coordinates of receiver A, and the northward coordinate difference is the difference between the northward coordinates of receiver B and the northward coordinates of receiver A. These two differences constitute the eastward and northward components of the baseline vector, reflecting the horizontal offset of the second patch antenna relative to the first patch antenna. In narrow passages, changes in the equipment's heading will directly alter the proportional relationship between these two components.

[0126] In detail, the baseline length is the square root of the sum of the squares of the eastward coordinate difference and the squares of the northward coordinate difference (i.e., the Pythagorean theorem), and its value is equal to the straight-line distance between the two antennas on the horizontal plane. The baseline length is an important indicator for verifying the validity of the solution. If the length deviates too much from the actual antenna spacing, it indicates that there is anomaly in the coordinate data (such as positioning jumps caused by obstruction).

[0127] In detail, the target angle is obtained by calculating the eastward and northward coordinate differences using a two-parameter arctangent function. This target angle is defined as the angle between the vectors formed by the eastward and northward coordinate differences, moving clockwise from north. This conforms to the conventional definition of heading angle in navigation (0 degrees for north, 90 degrees for east, 180 degrees for south, and 270 degrees for west). The two-parameter arctangent function can determine the quadrant of the vector by the sign of the coordinate difference, ensuring the uniqueness of the heading angle within the range of 0 to 360 degrees and avoiding the quadrant ambiguity problem of the single-parameter arctangent function.

[0128] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. An embedded AGV positioning and navigation method based on differential GNSS, characterized in that, include: Multiple passive global satellite navigation system patch antennas are evenly distributed along the circumference of the vehicle's roof; The radio frequency inputs of the two global satellite navigation system receivers are connected to any patch antenna, and the outputs are the carrier-to-noise ratio and elevation angle of each satellite; Determine the set of low-elevation satellites based on the elevation angles of all connected satellites; Calculate the carrier-to-noise ratio difference for a low-elevation-angle satellite array on two patch antennas symmetrically arranged on a circle. The environmental occlusion period is obtained by performing periodic analysis on the carrier-to-noise ratio difference, and the phase interval is divided. Within each phase interval, the first and second antennas corresponding to the low-elevation-angle satellite set are statistically analyzed. During the positioning and navigation operation period, the corresponding phase interval is determined to obtain the corresponding first antenna and second antenna; Real-time dynamic differential heading calculation is performed based on the first and second antennas.

2. The embedded AGV positioning and navigation method based on differential GNSS according to claim 1, characterized in that, The set of low-elevation satellites is determined based on the elevation angles of all connected satellites, including: At time ti, obtain the first set of satellites and the second set of satellites whose elevation angles are simultaneously output by the receivers of the two global satellite navigation systems. Calculate the intersection of the first set of satellites and the second set of satellites to obtain the common connected satellite set. Extract the elevation angles of all satellites in the common connected satellite set and sort them in ascending order to obtain the elevation angle sequence; The average elevation angle of the elevation angle sequence is used to obtain the segmentation threshold; If the elevation angle of the i-th satellite in the common connected satellite set is less than or equal to the segmentation threshold, it is marked as a low elevation angle satellite to form a low elevation angle satellite set.

3. The embedded AGV positioning and navigation method based on differential GNSS according to claim 2, characterized in that, For two patch antennas symmetrically arranged on a circle, calculate the carrier-to-noise ratio difference of a low-elevation satellite array, including: Determine any pair of symmetrical patch antennas to form an antenna pair; The carrier-to-noise ratio of each satellite in the low-elevation satellite array is synchronously acquired by RF multiplexing. Calculate the average carrier-to-noise ratio (CNR) of each patch antenna pair for each satellite in the antenna pair to obtain the first average CNR and the second average CNR. The difference between the first average carrier-to-noise ratio and the second average carrier-to-noise ratio is calculated to obtain the corresponding carrier-to-noise ratio difference.

4. The embedded AGV positioning and navigation method based on differential GNSS according to claim 3, characterized in that, Periodic analysis of the carrier-to-noise ratio difference yields the environmental occlusion period, including: Based on the carrier-to-noise ratio (CNR) difference of the antenna pair at N time points, a CNR difference sequence is formed; Calculate the mean of the carrier-to-noise ratio difference sequence to remove the mean from the carrier-to-noise ratio difference sequence, and obtain the mean-removed carrier-to-noise ratio difference sequence. If the i-th de-average load ratio difference in the de-average load ratio difference sequence is less than the (i-1)-th de-average load ratio difference, and the i-th de-average load ratio difference is less than the (i+1)-th de-average load ratio difference, then the i-th de-average load ratio difference is marked as a local minimum to form a set of local minima. If the i-th de-average load noise ratio difference in the de-average load noise ratio difference sequence is greater than the (i-1)-th de-average load noise ratio difference, and the i-th de-average load noise ratio difference is greater than the (i+1)-th de-average load noise ratio difference, then the i-th de-average load noise ratio difference is marked as a local maximum to form a set of local maxima. Calculate the autocorrelation function of the de-average noise ratio difference sequence, select the de-average noise ratio difference corresponding to the first local maximum, and obtain the environmental occlusion period.

5. The embedded AGV positioning and navigation method based on differential GNSS according to claim 4, characterized in that, Dividing the phase interval, including: Determine the position of each de-average load noise ratio difference in the local minimum set in the de-average load noise ratio difference sequence, and select the time corresponding to the de-average load noise ratio difference with the smallest position as the reference time. The time corresponding to each difference in the average load-to-noise ratio in the local minimum set is determined as the boundary time. Calculate the ratio of the difference between the boundary time and the reference time to the environmental occlusion period, and map the ratio to obtain the corresponding normalized time phase through the frac function; Arrange each normalized time phase in ascending order to obtain the normalized time phase sequence; Calculate the average value of adjacent normalized time phases in the normalized time phase sequence, and sort the average values ​​in ascending order to obtain the average value sequence; Extract the phase intervals corresponding to adjacent average values ​​in the average value sequence.

6. The embedded AGV positioning and navigation method based on differential GNSS according to claim 5, characterized in that, Within each phase interval, the first and second antennas corresponding to the low-elevation satellite set are statistically analyzed, including: For each phase interval in the phase interval set, determine each moment when the normalized time phase belongs to the phase interval, forming a time set; For each time u in the time set and each patch antenna k: the sum of the carrier-to-noise ratios of each satellite in the low-elevation satellite set at time u on the patch antenna k is divided by the number of satellites in the low-elevation satellite set to obtain the average carrier-to-noise ratio of the patch antenna k at the low-elevation angle at time u. The sum of the average carrier-to-noise ratio at low elevation angle for each time in the time set is divided by the number of time points in the time set to obtain the statistics of patch antenna k in the phase interval. Select the patch antenna with the largest statistic from all patch antennas as the first patch antenna corresponding to the phase interval; The patch antenna symmetrical to the first patch antenna is used as the second patch antenna.

7. The embedded AGV positioning and navigation method based on differential GNSS according to claim 6, characterized in that, Real-time dynamic differential heading calculation is performed based on the first and second antennas, including: Connect the first patch antenna to receiver A and the second patch antenna to receiver B; At the current moment, the same differential correction data is simultaneously injected into receiver A and receiver B, and receiver A and receiver B are forced to perform the calculation with the current moment as the time stamp; Receiver A outputs the current east and north coordinates, and receiver B outputs the current east and north coordinates. The eastward coordinate difference is calculated as the eastward coordinate of receiver B minus the eastward coordinate of receiver A, and the northward coordinate difference is calculated as the northward coordinate of receiver B minus the northward coordinate of receiver A. The baseline length is the square root of the sum of the squares of the eastward coordinate difference and the squares of the northward coordinate difference; The heading angle is the target angle calculated by using the two-parameter arctangent function on the eastward coordinate difference and the northward coordinate difference. The target angle is the angle between the vector formed by the northward clockwise direction and the vector composed of the eastward coordinate difference and the northward coordinate difference.

Images