A combined positioning solution method based on Beidou pseudo-range and 5G phase measurement

Abstract

This invention discloses a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement, relating to the field of BeiDou navigation and positioning technology. The method includes: Step 1: Generating a pseudorange direction compression domain; Step 2: Generating a phase continuous band set; Step 3: Generating a heterogeneous mutual attraction time series segment; Step 4: Inputting the heterogeneous mutual attraction time series segment into a heterogeneous phase reliable solver, with the string beam evolution unit setting a pseudorange phase string beam evolution mechanism to generate 5G phase reliable markers, phase break markers, and phase continuation markers; Step 5: Obtaining reliable 5G phase positioning constraints; Step 6: Constructing joint positioning solution constraints; Step 7: Obtaining joint positioning coordinates; Step 8: Obtaining the joint positioning trajectory and outputting the joint positioning solution result. This invention combines BeiDou pseudorange, 5G phase measurement, and a heterogeneous phase reliable solver to achieve highly stable joint positioning solution.

Description

Technical Field

[0001] This invention relates to the field of BeiDou navigation and positioning technology, and in particular to a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement. Background Technology

[0002] With the increasing demand for high-precision location services, low-altitude intelligent terminals, vehicle-road cooperative systems, and navigation in complex urban areas, multi-source joint positioning technology combining BeiDou pseudorange and 5G phase measurements has attracted widespread attention. Existing satellite navigation or cellular-assisted positioning methods mainly rely on single BeiDou pseudorange calculations, 5G base station ranging assistance, or direct combination of multi-source observations using least squares or Kalman filtering. However, these methods generally suffer from the following problems in practical applications: BeiDou pseudorange is prone to local anomalies in urban canyons, obstructed environments, and multipath environments. Existing methods typically include all visible satellite pseudoranges directly in the solution, lacking processing based on the spatial neighborhood relationship between BeiDou satellites and the terminal's predicted coordinates for continuous classification. This leads to abnormal pseudorange directions entering the positioning constraints and affecting the stability of terminal coordinate updates. 5G phase measurements have periodicity and integer discontinuity characteristics. Existing methods are insufficient in handling phase jumps in adjacent solution epochs, making it difficult to maintain continuous reconnection of single-base station phase differences through accumulated integer compensation, resulting in step breaks in the phase continuity band. BeiDou pseudorange and 5G phase have different sources and significantly different spatial line-of-sight directions. Existing joint solutions mostly remain at the level of residual weighting or filtering fusion, lacking spatial mutual attraction between the pseudorange direction compression domain and the phase continuity band set, as well as the evolutionary identification of the 5G phase reliability state, resulting in insufficient robustness and continuity of the final joint positioning solution constraints.

[0003] Therefore, how to provide a joint positioning solution based on BeiDou pseudorange and 5G phase measurement is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] One objective of this invention is to propose a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement. This invention combines BeiDou pseudorange, 5G phase measurement, and a heterogeneous phase reliable solver. It isolates abnormal directions in BeiDou pseudorange through a pseudorange direction compression domain, continuously reconnects 5G phase measurements through cumulative integer compensation, and performs recursive processing using a pseudorange phase bead evolution mechanism to generate 5G phase reliable markers, phase break markers, and phase continuation markers, and further obtains reliable 5G phase positioning constraints. Based on this, a joint positioning solution constraint is constructed, and the terminal predicted coordinates are iteratively updated, ultimately obtaining continuous, stable joint positioning coordinates and joint positioning trajectory with strong resistance to abnormal observations.

[0005] A joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to an embodiment of the present invention includes the following steps: Step 1: Calculate the pseudorange difference of a single satellite, and classify the pseudorange difference of a single satellite continuously according to the spatial neighborhood relationship of the Beidou satellite relative to the terminal's predicted coordinates, to generate a pseudorange direction compression domain; Step 2: Calculate the phase difference of a single base station, and continuously reconnect the phase difference of a single base station by accumulating integer compensation to generate a set of continuous phase bands; Step 3: Spatial cross-reference between the pseudorange direction compression domain and the phase continuous band set; based on the correspondence between the spatial direction of the 5G base station and the pseudorange direction compression domain, generate heterogeneous cross-reference timing segments. Step 4: Input the heterogeneous mutual reference timing segment into the heterogeneous phase trusted solver. The heterogeneous phase trusted solver includes a mutual reference coding unit, a phase consistency gating unit, a string beam evolution unit, and a trusted tag output unit. The string beam evolution unit sets up a pseudo-range phase string beam evolution mechanism to generate 5G phase trusted tags, phase break tags, and phase continuity tags. Step 5: Based on the 5G phase trust marker, phase break marker, and phase continuation marker, retain, delete, and correct the direction consistency of the phase continuous band set to obtain the trustworthy 5G phase positioning constraint; Step 6: Obtain the BeiDou pseudorange direction constraint based on the pseudorange direction compression domain, obtain the 5G phase continuity constraint based on the reliable 5G phase positioning constraint, and construct the joint positioning solution constraint; Step 7: Update the terminal predicted coordinates based on the joint positioning solution constraints, and iteratively generate updated joint positioning solution constraints until the coordinate change does not exceed the set coordinate change threshold, and obtain the joint positioning coordinates; Step 8: Connect the joint positioning coordinates according to the solution epoch order to obtain the joint positioning trajectory, and output the joint positioning solution result.

[0006] Optionally, step one specifically includes: Obtain the BeiDou pseudorange, BeiDou satellite position, and terminal predicted coordinates within the same solution epoch, and match the BeiDou pseudorange with the BeiDou satellite position according to the BeiDou satellite number; The straight-line distance between the terminal's predicted coordinates and the positions of each BeiDou satellite is calculated to obtain the single-satellite predicted geometric distance. The BeiDou pseudorange corresponding to each BeiDou satellite is then subtracted from the corresponding single-satellite predicted geometric distance to obtain the single-satellite pseudorange difference. Obtain the spatial line-of-sight vector of each BeiDou satellite position relative to the terminal's predicted coordinates, and normalize the spatial line-of-sight vector to obtain the single-satellite spatial unit direction vector; Map the spatial unit direction vector of each single satellite to the local celestial coordinate system with the terminal predicted coordinates as the origin, and calculate the spatial solid angle between the spatial unit direction vectors of any two Beidou satellites. BeiDou satellites whose spatial three-dimensional angle does not exceed the set neighborhood angle are classified into the same adjacent satellite group; The pseudorange difference values ​​of individual BeiDou satellites within the same adjacent satellite group are marked with positive and negative signs to obtain satellites with non-negative difference values ​​and satellites with negative difference values. When the number of BeiDou satellites in the same adjacent satellite group is greater than the set base threshold, and the number of non-negative difference satellites or negative difference satellites in the group is greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is determined as a directional continuous adjacent group. When the number of BeiDou satellites in an adjacent satellite group is not greater than the set base threshold, or when the number of non-negative difference satellites and the number of negative difference satellites in the group are both not greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is defined as a directionally dispersed adjacent group. The single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of continuous adjacency group are classified into the pseudorange contraction direction region, and the single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of dispersed adjacency group are classified into the pseudorange repulsion direction region. The pseudorange contraction direction region and the pseudorange repulsion direction region are bound together according to the same solution epoch to generate a pseudorange direction compression domain.

[0007] Optionally, step two specifically involves: Obtain 5G phase measurement values, 5G base station locations, 5G signal wavelengths, and terminal predicted coordinates within the same solution epoch; Multiply the 5G phase measurement value by the corresponding 5G signal wavelength to obtain the initial phase distance of a single base station; Calculate the straight-line distance between the terminal's predicted coordinates and the location of each 5G base station to obtain the predicted geometric distance of a single base station, and subtract the predicted geometric distance of the corresponding single base station from the initial phase distance of each 5G base station to obtain the phase difference of the single base station. Arrange the single base station phase difference values ​​corresponding to the same 5G base station in the order of the solution epochs, set the cumulative integer compensation amount of the same 5G base station in the first solution epoch to zero, and take the single base station phase difference value in the first solution epoch as the continuous reconnection phase difference value in the first solution epoch. Starting from the second epoch involved in the solution, the difference between the single-base station phase difference of the current solution epoch and the single-base station phase difference of the adjacent previous solution epoch is calculated to obtain the phase evolution difference between epochs. When the phase evolution difference between epochs is greater than the set positive cycle jump threshold, the cumulative integer compensation amount corresponding to the previous solution epoch is increased by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is less than the set reverse cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is reduced by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is between the set reverse cycle slip threshold and the set forward cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is used as the cumulative integer compensation amount corresponding to the current solution epoch. The single base station phase difference value of the current solution epoch is deducted and corrected based on the cumulative integer compensation amount corresponding to the current solution epoch to obtain the continuous reconnection phase difference value. The continuous phase difference values ​​of the same 5G base station in the continuous solution epoch are connected in the order of solution epoch to obtain the continuous phase band of the base station, and the continuous phase bands of each base station are summarized to generate a continuous phase band set.

[0008] Optionally, step three specifically includes: Subtract the terminal's predicted coordinates from the 5G base station location within the same solution epoch to obtain the base station spatial line-of-sight vector. Then, normalize the base station spatial line-of-sight vector to generate the 5G base station spatial unit direction vector. Calculate the spatial solid angle between the spatial unit direction vector of the 5G base station and the spatial unit direction vector of each single star in the pseudorange direction compression domain to obtain the set of base station star direction angles. Select the smallest spatial three-dimensional angle from the set of base station star orientation angles, and determine the single-star spatial unit direction vector corresponding to the smallest spatial three-dimensional angle as the mutual attraction BeiDou direction of the 5G base station, and determine the smallest spatial three-dimensional angle as the mutual attraction adaptation angle. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange contraction direction area in the pseudorange direction compression domain, the 5G base station is marked as a contraction mutual attraction base station. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange exclusion direction region in the pseudorange direction compression domain, the 5G base station will be marked as a mutually exclusive base station. When the mutual referencing and adaptation angle exceeds the set spatial mutual referencing angle threshold, the 5G base station will be marked as an independent mutual referencing base station. According to the 5G base station number, the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations, and independent mutual referencing base stations are written into the base station mutual referencing index table; Extract the phase continuous bands of each base station corresponding to the base station mutual reference index table from the phase continuous band set, and calculate the change in the continuous reconnection phase difference between the current solution epoch and the adjacent previous solution epoch in each base station phase continuous band. The base station mutual reference index table and the change in phase difference of continuous reconnection are spatiotemporally bound according to the solution epoch to generate heterogeneous mutual reference time sequence segments.

[0009] Optionally, step four specifically includes: The heterogeneous mutual reference time sequence segment is input into the mutual reference encoding unit in the heterogeneous phase reliable solver. The base station mutual reference index table and the change in continuous reconnection phase difference in the heterogeneous mutual reference time sequence segment are vectorized and arranged to obtain the mutual reference input vector. The mutual reference input vector is input into a one-dimensional convolutional layer for encoding, and then non-linearly activated by the GELU activation function to obtain the mutual reference encoded features; The mutual referencing coding features are input into the phase consistency gating unit, and mutual referencing type coding is generated for each 5G base station based on the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations and independent mutual referencing base stations in the base station mutual referencing index table. The mutual reference type code and the corresponding mutual reference code feature of the 5G base station are concatenated according to the channel dimension to obtain the mutual reference type concatenation feature; The concatenated features of mutual reference types are input into the gated mapping layer and phase consistency gating coefficients are generated by the Sigmoid activation function. The phase consistency gating coefficients are multiplied element-wise with the cross-reference coding features to obtain the phase consistency gating features; The phase consistency gating features are input into the string beam evolution unit, and the pseudorange phase string beam evolution mechanism is executed to perform evolution processing on the phase consistency gating features in adjacent solution epochs to obtain the single base station chain evolution features of each 5G base station. The single-base station chain evolution feature of each 5G base station is input into the trusted label output unit. Through the linear mapping layer, the evolution state mapping vector is obtained. The evolution state mapping vector is then input into the Softmax function to obtain the continuous progressive state probability and interruption state probability of the corresponding 5G base station. When the probability of an interrupted state is greater than the probability of a continuous progressive state, a phase break marker is generated for the corresponding 5G base station; when the probability of a continuous progressive state is greater than or equal to the probability of an interrupted state, a phase continuation marker is generated for the corresponding 5G base station. 5G base stations with phase continuity markings and corresponding phase consistency gating coefficients greater than the set gating threshold are marked as 5G phase trust markings.

[0010] Optionally, the pseudorange phase bead recursion mechanism performs recursive processing on the phase consistency gating features in adjacent solution epochs, specifically as follows: The phase consistency gating features of the same 5G base station in each solution epoch are obtained and arranged in the order of solution epochs to generate a single base station phase gating feature sequence. Set the phase-gated differential feature corresponding to the first epoch to zero vector. Starting from the second epoch, subtract the phase-consistency gated feature corresponding to the adjacent previous epoch from the phase-consistency gated feature corresponding to the current epoch to obtain the phase-gated differential feature corresponding to the current epoch. The phase consistency gating feature and the corresponding phase gating differential feature corresponding to each solution epoch are concatenated according to the channel dimension to obtain the temporal string unrolling feature; The temporal beam unpacking features are sequentially input into the first linear mapping layer, the LayerNorm layer, the GELU activation layer, and the second linear mapping layer to obtain the static beam evolution features of a single base station. The chain-beam evolution feature output by the same 5G base station in the previous adjacent solution epoch is taken as the historical state. The historical state is superimposed with the static beam evolution feature of the single base station corresponding to the current solution epoch to generate the chain-beam evolution feature of the current solution epoch. The historical state corresponding to the first solution epoch is initialized as a zero vector. According to the order of the solution epochs, the chain-like chord progression feature of the current solution epoch is used as the historical state of the next solution epoch and is rolled over until the last solution epoch generates the chain-like chord progression feature, which is then used as the single-base station chain progression feature of the corresponding 5G base station.

[0011] Optionally, step five specifically includes: The continuous band of the base station phase corresponding to the 5G base station with 5G phase trust label is retained to obtain the trust phase continuous band; Delete the continuous phase band of the 5G base station corresponding to the phase break marker; The phase continuity band of the 5G base station with phase continuity mark but without 5G phase trust mark is taken as the phase continuity band to be corrected and directional consistency correction is performed. The number of positive and negative directions of the change in the continuous reconnection phase difference of each reliable phase continuous band within the same solution epoch is counted, and the direction with the larger number is determined as the reference change direction; The continuous back-loop phase difference value of the first participating epoch in the phase continuum to be corrected is taken as the corrected continuous back-loop phase difference value corresponding to the first participating epoch. Starting from the second participating epoch, the continuous back-loop phase difference value of the current solving epoch in the phase continuum to be corrected is subtracted from the continuous back-loop phase difference value of the adjacent previous solving epoch to obtain the phase change value to be corrected corresponding to the current solving epoch. When the positive and negative directions of the phase change to be corrected are consistent with the reference change direction, the continuous loop phase difference of the current solution epoch is taken as the corrected continuous loop phase difference of the current solution epoch. When the positive and negative directions of the phase change to be corrected are inconsistent with the reference change direction, the corrected continuous loop phase difference value corresponding to the adjacent previous solution epoch is used as the recursive reference. If the reference change direction is positive, the absolute value of the phase change to be corrected is added to the recursive reference. If the reference change direction is negative, the absolute value of the phase change to be corrected is subtracted from the recursive reference to obtain the corrected continuous loop phase difference value corresponding to the current solution epoch. By connecting the corrected phase difference values ​​in the order of the solution epochs, a corrected phase continuous band is obtained. The trusted phase continuous band and the modified phase continuous band are combined according to the 5G base station number to obtain the trusted 5G phase positioning constraint.

[0012] Optionally, step six specifically includes: Extract the single-satellite spatial unit direction vector corresponding to the pseudorange contraction direction region from the pseudorange direction compression domain, and extract the corresponding single-satellite pseudorange difference as the BeiDou pseudorange residual, which together serve as the BeiDou pseudorange direction constraint. The continuous back-up phase difference corresponding to the trusted phase continuous band and the modified phase continuous band is extracted from the trusted 5G phase positioning constraint and used as the 5G phase continuity constraint. The 5G base station numbers participating in the solution are determined based on the 5G phase continuity constraints, and the 5G base station locations corresponding to the 5G base station numbers participating in the solution are obtained. The straight-line distances between the terminal's predicted coordinates and the corresponding 5G base station locations are calculated to obtain the base station predicted distances. Subtract the predicted distance of the corresponding base station from the continuous reconnection phase difference value of the 5G base station to obtain the continuous phase residual of 5G. A spatial geometric matrix is ​​established based on the single-star spatial unit direction vector in the pseudorange direction constraint of Beidou, without introducing the single-star spatial unit direction vector corresponding to the pseudorange exclusion direction region, thus limiting the direction update range of the terminal prediction coordinates. The spatial geometric matrix, BeiDou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and corresponding solution epoch are bound together to generate single epoch joint constraint terms; Arrange the joint constraint terms of each single epoch in the order of solution epochs to generate joint positioning solution constraints.

[0013] Optionally, step seven specifically includes: Read the single-epoch joint constraint terms of the corresponding solution epoch one by one from the joint positioning solution constraints, and extract the spatial geometric matrix, Beidou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and terminal predicted coordinates. Based on the single-star spatial unit direction vector in the spatial geometric matrix and the corresponding BeiDou pseudorange residual, the BeiDou coordinate correction is generated. The location of the corresponding 5G base station is obtained based on the corresponding 5G base station number, and the 5G coordinate correction amount is generated based on the spatial direction of the 5G base station location relative to the terminal's predicted coordinates and the 5G phase continuity residual. The BeiDou coordinate correction and the 5G coordinate correction are added together according to the horizontal coordinate, vertical coordinate and elevation coordinate to obtain the joint coordinate correction. The joint coordinate correction is superimposed on the terminal predicted coordinates according to the horizontal coordinate, vertical coordinate and elevation coordinate respectively, and the terminal predicted coordinates are updated to obtain the updated terminal predicted coordinates. The updated joint positioning solution constraints are then regenerated based on the updated terminal predicted coordinates. Calculate the coordinate change between the updated terminal predicted coordinates and the terminal predicted coordinates before the update; When the coordinate change exceeds the set coordinate change threshold, the terminal predicted coordinates continue to be updated based on the updated joint positioning solution constraints. When the coordinate change does not exceed the set coordinate change threshold, the updated terminal predicted coordinates will be determined as the joint positioning coordinates.

[0014] Optionally, step eight specifically includes: According to the order of solution epochs, obtain the joint positioning coordinates obtained by iterative solution in each solution epoch; Connect the joint positioning coordinates corresponding to adjacent solution epochs in chronological order to generate a joint positioning trajectory; Calculate the coordinate difference between two adjacent joint positioning coordinates to obtain the displacement of the trajectory segment; The joint positioning coordinates, joint positioning trajectory, trajectory segment displacement, and corresponding solution epochs are encapsulated and output as joint positioning solution results.

[0015] The beneficial effects of this invention are: This invention addresses the issue of abnormal pseudorange direction interference in BeiDou pseudorange under occlusion and multipath environments by constructing a cascaded pseudorange direction compression domain, a phase continuous band set, and a heterogeneous phase reliable solver. It classifies the pseudorange difference of individual satellites according to the spatial neighborhood relationship between the BeiDou satellite and the terminal's predicted coordinates. The single-satellite spatial unit direction vector and pseudorange difference corresponding to the directionally continuous adjacent group are classified into the pseudorange contraction direction region, while the single-satellite spatial unit direction vector and pseudorange difference corresponding to the directionally dispersed adjacent group are classified into the pseudorange exclusion direction region, achieving effective screening of BeiDou pseudorange direction constraints. To address the integer discontinuity and step jump problems in 5G phase measurement, the phase difference of a single base station is deducted and corrected by accumulating integer compensation, generating a phase continuous band set with temporal continuity. Finally, to address the spatial direction inconsistency between BeiDou pseudorange and 5G phase, this invention... The compressed domain and the phase continuum set are spatially mutually referenced to generate heterogeneous mutual reference time segments, which are then input into the heterogeneous phase reliable solver. Through the mutual reference coding unit, phase consistency gating unit, string beam evolution unit, and reliable tag output unit, 5G phase reliable tags, phase break tags, and phase continuation tags are generated. Among them, the string beam evolution unit sets up a pseudorange phase string beam evolution mechanism, which can perform chain evolution processing on the phase consistency gating features in adjacent solution epochs, so that the phase continuity state of each 5G base station is continuously transmitted on the time axis, thereby improving the stability of 5G phase reliable tags in identifying phase continuation and phase break. Subsequently, based on the reliable 5G phase positioning constraints and Beidou pseudorange direction constraints, a joint positioning solution constraint is constructed, and the terminal predicted coordinates are iteratively updated to obtain joint positioning coordinates and joint positioning trajectory, which improves the stability, continuity, and anti-anomaly observation capability of joint positioning solution in complex environments. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is an overall flowchart of a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement proposed in this invention. Figure 2 This is a schematic diagram of the heterogeneous phase reliable solver for a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement proposed in this invention. Figure 3 This is a schematic diagram of the process for generating a joint positioning trajectory using a joint positioning solution method based on BeiDou pseudorange and 5G phase measurement proposed in this invention. Detailed Implementation

[0017] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0018] refer to Figures 1-3 A joint positioning solution method based on BeiDou pseudorange and 5G phase measurement includes the following steps: Step 1: Calculate the pseudorange difference of a single satellite, and classify the pseudorange difference of a single satellite continuously according to the spatial neighborhood relationship of the Beidou satellite relative to the terminal's predicted coordinates, to generate a pseudorange direction compression domain; Step 2: Calculate the phase difference of a single base station, and continuously reconnect the phase difference of a single base station by accumulating integer compensation to generate a set of continuous phase bands; Step 3: Spatial cross-reference between the pseudorange direction compression domain and the phase continuous band set; based on the correspondence between the spatial direction of the 5G base station and the pseudorange direction compression domain, generate heterogeneous cross-reference timing segments. Step 4: Input the heterogeneous mutual reference timing segment into the heterogeneous phase trusted solver. The heterogeneous phase trusted solver includes a mutual reference coding unit, a phase consistency gating unit, a string beam evolution unit, and a trusted tag output unit. The string beam evolution unit sets up a pseudo-range phase string beam evolution mechanism to generate 5G phase trusted tags, phase break tags, and phase continuity tags. Step 5: Based on the 5G phase trust marker, phase break marker, and phase continuation marker, retain, delete, and correct the direction consistency of the phase continuous band set to obtain the trustworthy 5G phase positioning constraint; Step 6: Obtain the BeiDou pseudorange direction constraint based on the pseudorange direction compression domain, obtain the 5G phase continuity constraint based on the reliable 5G phase positioning constraint, and construct the joint positioning solution constraint; Step 7: Update the terminal predicted coordinates based on the joint positioning solution constraints, and iteratively generate updated joint positioning solution constraints until the coordinate change does not exceed the set coordinate change threshold, and obtain the joint positioning coordinates; Step 8: Connect the joint positioning coordinates according to the solution epoch order to obtain the joint positioning trajectory, and output the joint positioning solution result.

[0019] In this embodiment, step one specifically includes: Obtain the BeiDou pseudorange, BeiDou satellite position, and terminal predicted coordinates within the same solution epoch, and match the BeiDou pseudorange with the BeiDou satellite position according to the BeiDou satellite number; The straight-line distance between the terminal's predicted coordinates and the positions of each BeiDou satellite is calculated to obtain the single-satellite predicted geometric distance. The BeiDou pseudorange corresponding to each BeiDou satellite is then subtracted from the corresponding single-satellite predicted geometric distance to obtain the single-satellite pseudorange difference. Obtain the spatial line-of-sight vector of each BeiDou satellite position relative to the terminal's predicted coordinates, and normalize the spatial line-of-sight vector to obtain the single-satellite spatial unit direction vector; Map the spatial unit direction vector of each single satellite to the local celestial coordinate system with the terminal predicted coordinates as the origin, and calculate the spatial solid angle between the spatial unit direction vectors of any two Beidou satellites. BeiDou satellites whose spatial three-dimensional angle does not exceed the set neighborhood angle are classified into the same adjacent satellite group; The pseudorange difference values ​​of individual BeiDou satellites within the same adjacent satellite group are marked with positive and negative signs to obtain satellites with non-negative difference values ​​and satellites with negative difference values. When the number of BeiDou satellites in the same adjacent satellite group is greater than the set base threshold, and the number of non-negative difference satellites or negative difference satellites in the group is greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is determined as a directional continuous adjacent group. When the number of BeiDou satellites in an adjacent satellite group is not greater than the set base threshold, or when the number of non-negative difference satellites and the number of negative difference satellites in the group are both not greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is defined as a directionally dispersed adjacent group. The single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of continuous adjacency group are classified into the pseudorange contraction direction region, and the single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of dispersed adjacency group are classified into the pseudorange repulsion direction region. The pseudorange contraction direction region and the pseudorange repulsion direction region are bound together according to the same solution epoch to generate a pseudorange direction compression domain. In the specific implementation process, the same calculation epoch is set to 1 second. The terminal predicted coordinates and BeiDou satellite positions are uniformly converted to the geocentric coordinate system, with the coordinate unit being meters. BeiDou pseudoranges are matched one by one according to the BeiDou satellite number to avoid mismatch between observation values ​​and satellite positions. The spatial line-of-sight vector is obtained by subtracting the terminal predicted coordinates from the BeiDou satellite position. During the unitization process, the three coordinate components of the spatial line-of-sight vector are divided by the length of the spatial line-of-sight vector, so that the direction vector of a single satellite spatial unit retains only the direction information. The local celestial coordinate system takes the terminal predicted coordinates as the origin and uses the local east, local north, and celestial directions as the three coordinate axes. The spatial solid angle is determined according to the angle between the direction vectors of two single satellite spatial units. The neighborhood angle is set to 30 degrees, the proportion is set to two-thirds, and the base threshold is set to 2 satellites. When the number of satellites in the group is less than 2, since effective spatial collaborative verification cannot be formed, it is directly judged as a directionally dispersed adjacent group to ensure the focus of computing power and the absolute reliability of the correction direction.

[0020] When multiple BeiDou satellites form an adjacent satellite group within the spatial three-dimensional angle range, it indicates that these BeiDou satellites have similar spatial observation directions around the terminal predicted coordinates. If the number of non-negative difference satellites or negative difference satellites in the adjacent satellite group is greater than two-thirds, it means that the BeiDou pseudorange relative to the single-satellite predicted geometric distance in this spatial neighborhood shows a consistent deviation direction. The corresponding single-satellite spatial unit direction vector and single-satellite pseudorange difference are classified into the pseudorange contraction direction region to retain stable direction correction information. If the number of non-negative difference satellites and negative difference satellites are both less than two-thirds, it means that the deviation direction in this spatial neighborhood is dispersed. The corresponding content is classified into the pseudorange exclusion direction region to isolate the interference of local multipath, obstruction, or observational abrupt changes on the pseudorange direction compression domain, thereby improving the stability of spatial mutual attraction and joint positioning solution constraints.

[0021] In this embodiment, step two specifically includes: Obtain 5G phase measurement values, 5G base station locations, 5G signal wavelengths, and terminal predicted coordinates within the same solution epoch; Multiply the 5G phase measurement value by the corresponding 5G signal wavelength to obtain the initial phase distance of a single base station; Calculate the straight-line distance between the terminal's predicted coordinates and the location of each 5G base station to obtain the predicted geometric distance of a single base station, and subtract the predicted geometric distance of the corresponding single base station from the initial phase distance of each 5G base station to obtain the phase difference of the single base station. Arrange the single base station phase difference values ​​corresponding to the same 5G base station in the order of the solution epochs, set the cumulative integer compensation amount of the same 5G base station in the first solution epoch to zero, and take the single base station phase difference value in the first solution epoch as the continuous reconnection phase difference value in the first solution epoch. Starting from the second epoch involved in the solution, the difference between the single-base station phase difference of the current solution epoch and the single-base station phase difference of the adjacent previous solution epoch is calculated to obtain the phase evolution difference between epochs. When the phase evolution difference between epochs is greater than the set positive cycle jump threshold, the cumulative integer compensation amount corresponding to the previous solution epoch is increased by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is less than the set reverse cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is reduced by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is between the set reverse cycle slip threshold and the set forward cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is used as the cumulative integer compensation amount corresponding to the current solution epoch. The single base station phase difference value of the current solution epoch is deducted and corrected based on the cumulative integer compensation amount corresponding to the current solution epoch to obtain the continuous reconnection phase difference value. The continuous phase difference values ​​of the same 5G base station in the continuous solution epoch are connected in the order of the solution epoch to obtain the continuous phase band of the base station, and the continuous phase bands of each base station are summarized to generate a continuous phase band set. In the specific implementation process, 5G phase measurements are expressed as phase observations in terms of carrier cycles. The 5G signal wavelength is 0.01 meters, and the same calculation epoch is 1 second. The 5G base station location and the terminal predicted coordinates are uniformly transformed to the same geocentric geofixed coordinate system. The initial phase distance of a single base station is used to represent the distance-based observation result corresponding to the 5G phase measurement value. The predicted geometric distance of a single base station is used to represent the current estimated distance from the terminal predicted coordinates to the corresponding 5G base station location. The phase difference of a single base station is used to represent the deviation between the distance-based phase observation result and the current estimated distance. Since 5G phase measurements have integer cycle discontinuities, the phase difference of a single base station directly connecting each calculation epoch will produce a step transition. Therefore, the cumulative integer cycle compensation is set to zero in the first calculation epoch, serving as the recursive starting point for subsequent continuous reconnection.

[0022] Starting from the second participating epoch, the phase evolution difference between epochs is used to determine whether the current epoch has a cycle jump relative to the adjacent previous epoch. The positive cycle jump threshold is set to 0.006 meters, and the negative cycle jump threshold is set to -0.006 meters. The deduction correction refers to subtracting the cumulative cycle compensation amount corresponding to the current epoch from the single-base station phase difference value of the current epoch to obtain the continuous reconnection phase difference value. When the cumulative cycle compensation amount is positive, it indicates that a positive cycle jump has occurred in the previous epoch, and the corresponding cycle distance needs to be deducted from the current single-base station phase difference value. When the cumulative cycle compensation amount is negative, it indicates that a negative cycle jump has occurred in the previous epoch, and the current single-base station phase difference value minus the negative cumulative cycle compensation amount is equivalent to replenishing the corresponding cycle distance. Through this deduction correction, the step offset formed by a single cycle jump can continue to act on the phase reconnection process of subsequent epochs, avoiding only correcting the current epoch and causing a break again in the next epoch. The base station phase continuity band formed by the continuous back-connection phase difference connection has time continuity, which can reduce the interference of periodic jumps and local phase breaks in 5G phase measurement on the phase continuity band set, and provide a stable 5G phase continuity basis for the generation of heterogeneous mutual reference timing segments.

[0023] In this embodiment, step three specifically includes: Subtract the terminal's predicted coordinates from the 5G base station location within the same solution epoch to obtain the base station spatial line-of-sight vector. Then, normalize the base station spatial line-of-sight vector to generate the 5G base station spatial unit direction vector. Calculate the spatial solid angle between the spatial unit direction vector of the 5G base station and the spatial unit direction vector of each single star in the pseudorange direction compression domain to obtain the set of base station star direction angles. Select the smallest spatial three-dimensional angle from the set of base station star orientation angles, and determine the single-star spatial unit direction vector corresponding to the smallest spatial three-dimensional angle as the mutual attraction BeiDou direction of the 5G base station, and determine the smallest spatial three-dimensional angle as the mutual attraction adaptation angle. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange contraction direction area in the pseudorange direction compression domain, the 5G base station is marked as a contraction mutual attraction base station. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange exclusion direction region in the pseudorange direction compression domain, the 5G base station will be marked as a mutually exclusive base station. When the mutual referencing and adaptation angle exceeds the set spatial mutual referencing angle threshold, the 5G base station will be marked as an independent mutual referencing base station. According to the 5G base station number, the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations, and independent mutual referencing base stations are written into the base station mutual referencing index table; Extract the phase continuous bands of each base station corresponding to the base station mutual reference index table from the phase continuous band set, and calculate the change in the continuous reconnection phase difference between the current solution epoch and the adjacent previous solution epoch in each base station phase continuous band. The base station mutual reference index table and the change in continuous back-connection phase difference are spatiotemporally bound according to the solution epoch to generate heterogeneous mutual reference timing segments. In the specific implementation process, the 5G base station location and the terminal predicted coordinates are uniformly adopted in the geocentric and geofixed coordinate system, with the coordinate unit being meters. The base station spatial line-of-sight vector is obtained by subtracting the terminal predicted coordinates from the 5G base station location. When normalizing the base station spatial line-of-sight vector, the three coordinate components of the base station spatial line-of-sight vector are divided by the length of the base station spatial line-of-sight vector to obtain a 5G base station spatial unit direction vector with a length of 1. This ensures that the 5G base station spatial unit direction vector only expresses the spatial orientation of the 5G base station relative to the terminal predicted coordinates. The spatial mutual attraction angle threshold is set at 45 degrees. When the mutual attraction adaptation angle exceeds 45 degrees, it indicates that the 5G base station is not in a spatial direction that can be effectively pulled by the pseudorange direction compression domain. Therefore, it is marked as an independent mutual attraction base station to avoid forcibly classifying or excluding 5G base stations with weak spatial correlation into shrinking mutual attraction base stations.

[0024] The base station mutual referencing index table records the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations, or independent mutual referencing base stations corresponding to each 5G base station number, enabling the base station phase continuum in the phase continuum set to be reorganized according to spatial mutual referencing attributes. The change in continuous loop phase difference is obtained by subtracting the continuous loop phase difference of the adjacent previous solution epoch from the continuous loop phase difference of the current solution epoch, and is used to characterize the phase change trend of the corresponding 5G base station in the phase continuum. After the base station mutual referencing index table and the change in continuous loop phase difference are spatiotemporally bound according to the solution epoch, the heterogeneous mutual referencing time sequence segment simultaneously carries the spatial mutual referencing category and phase continuum change information of the 5G base station, providing structured input for the heterogeneous phase trusted solver to identify 5G phase trusted markers, phase break markers, and phase continuation markers, thereby improving the reliability of the screening of trusted 5G phase positioning constraints.

[0025] In this embodiment, step four specifically includes: The heterogeneous mutual reference time sequence segment is input into the mutual reference encoding unit in the heterogeneous phase reliable solver. The base station mutual reference index table and the change in continuous reconnection phase difference in the heterogeneous mutual reference time sequence segment are vectorized and arranged to obtain the mutual reference input vector. The mutual reference input vector is input into a one-dimensional convolutional layer for encoding, and then non-linearly activated by the GELU activation function to obtain the mutual reference encoded features; The mutual referencing coding features are input into the phase consistency gating unit, and mutual referencing type coding is generated for each 5G base station based on the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations and independent mutual referencing base stations in the base station mutual referencing index table. The mutual reference type code and the corresponding mutual reference code feature of the 5G base station are concatenated according to the channel dimension to obtain the mutual reference type concatenation feature; The concatenated features of mutual reference types are input into the gated mapping layer and phase consistency gating coefficients are generated by the Sigmoid activation function. The phase consistency gating coefficients are multiplied element-wise with the cross-reference coding features to obtain the phase consistency gating features; The phase consistency gating features are input into the string beam evolution unit, and the pseudorange phase string beam evolution mechanism is executed to perform evolution processing on the phase consistency gating features in adjacent solution epochs to obtain the single base station chain evolution features of each 5G base station. The single-base station chain evolution feature of each 5G base station is input into the trusted label output unit. Through the linear mapping layer, the evolution state mapping vector is obtained. The evolution state mapping vector is then input into the Softmax function to obtain the continuous progressive state probability and interruption state probability of the corresponding 5G base station. When the probability of an interrupted state is greater than the probability of a continuous progressive state, a phase break marker is generated for the corresponding 5G base station; when the probability of a continuous progressive state is greater than or equal to the probability of an interrupted state, a phase continuation marker is generated for the corresponding 5G base station. 5G base stations with phase continuity markings and corresponding phase consistency gating coefficients greater than a set gating threshold are marked as 5G phase trust markings; In the specific implementation process, the heterogeneous mutual referencing time series segments are organized into a two-dimensional time series table according to the solution epoch and 5G base station number. Each row corresponds to the base station mutual referencing index table marker field and the continuous reconnection phase difference change amount of a 5G base station under a solution epoch. When vectorizing the arrangement, the base station mutual referencing index table marker field is retained as an identifiable category field, and the continuous reconnection phase difference change amount is processed by maximum and minimum normalization and then arranged into the phase change channel to obtain the mutual referencing input vector. The kernel length of the one-dimensional convolutional layer is set to 3 and the stride is set to 1 to extract the local evolution features of the continuous reconnection phase difference change amount within three adjacent solution epochs. The GELU activation function is used to retain the continuous trend in small phase changes and reduce the feature loss caused by hard truncation.

[0026] In the phase consistency gating unit, based on the marker field of the base station mutual referencing index table, the shrinking mutual referencing base station is converted into a three-dimensional mutual referencing type code with the first bit set to 1, the second bit to 0, and the third bit to 0; the repulsive mutual referencing base station is converted into a three-dimensional mutual referencing type code with the first bit set to 0, the second bit to 1, and the third bit to 0; and the independent mutual referencing base station is converted into a three-dimensional mutual referencing type code with the first bit set to 0, the second bit to 0, and the third bit to 1. Then, the mutual referencing type code is concatenated with the mutual referencing code feature of the corresponding 5G base station according to the channel dimension to obtain the mutual referencing type concatenation feature. The gating mapping layer adopts a fully connected mapping structure, compressing the mutual referencing type concatenation feature into a single gating output value, and mapping it to between 0 and 1 using the Sigmoid activation function to obtain the phase consistency gating coefficient of the 5G base station in the current solution epoch. The phase consistency gating coefficient is used to characterize the corresponding 5G base station. The consistency between spatial mutual attraction type and continuous phase change is assessed. When the phase consistency gating coefficient is close to 1, the corresponding mutual attraction coding feature is retained to a high degree in subsequent evolutionary processing. When the phase consistency gating coefficient is close to 0, the corresponding mutual attraction coding feature is suppressed in subsequent evolutionary processing. The gating threshold is set to 0.65. When the phase consistency gating coefficient is greater than 0.65, it indicates that the spatial mutual attraction attribute of the corresponding 5G base station and the change in the phase difference of continuous reconnection have phase consistency that can participate in the evolution. The linear mapping layer in the trusted label output unit maps the chain evolution feature of a single base station to two state channels. The Softmax function outputs the continuous progressive state probability and the interrupted state probability, respectively. When the two state probabilities are equal, they are classified into the continuous progressive state, which can avoid excessive generation of phase break tags in the critical state, thereby improving the stability of 5G phase trusted labels.

[0027] The training sample dataset for the heterogeneous phase reliable solver is constructed from BeiDou pseudorange, BeiDou satellite position, 5G phase measurement value, 5G base station position, 5G signal wavelength, and terminal predicted coordinates under continuous solution epochs. Specifically, the construction process involves first calculating the pseudorange difference of a single satellite according to step one of this invention, and forming a pseudorange direction compression domain based on the spatial neighborhood relationship of the BeiDou satellite relative to the terminal predicted coordinates and the sign consistency of the pseudorange difference of the single satellite; then, according to step two, performing cumulative integer compensation and continuous reconnection on the phase difference of a single base station to form a phase continuous band set; subsequently, according to step three, spatially interreferencing the pseudorange direction compression domain and the phase continuous band set to form a heterogeneous interreferencing time sequence segment. Each training sample uses the base station interreferencing type, continuous reconnection phase difference change, phase consistency gating information, and adjacent epoch difference information of the same 5G base station within multiple consecutive solution epochs as sample features, and the phase continuity state or phase break state of the 5G base station in the target solution epoch as sample labels. The sample sources include measured trajectory data, playback data from complex urban areas outdoors, and simulated phase jump data constructed based on real base station locations, real satellite ephemeris, and set terminal motion trajectories. Among them, measured data is used to ensure that the training samples are close to the real positioning environment, while simulated phase jump data is used to supplement sample types such as phase breaks, integer jumps, and sudden occlusions, which occur less frequently in the actual measurement process but have a greater impact on positioning stability.

[0028] The phase continuity and phase break markers are not arbitrary labels, but are determined based on the changes in the continuous reconnection phase difference of the 5G phase continuity band between adjacent solution epochs, the changes in the cumulative integer compensation, and the spatial consistency between the phase change direction of the 5G base station and the credible BeiDou pseudorange direction constraint. If the continuous reconnection phase difference of the same 5G base station changes smoothly in adjacent solution epochs, and the direction of change is consistent with the reference change direction of the credible phase continuity band within the same solution epoch, then the sample is marked as phase continuity. If the continuous reconnection phase difference of the same 5G base station changes abruptly in adjacent solution epochs, or if the cumulative integer compensation changes abnormally, and the direction of the phase change after the abrupt change is inconsistent with the reference change direction, then the sample is marked as phase break. For samples where phase changes are at the boundary, the residual changes between the terminal's predicted coordinates and the high-precision reference trajectory can be used for verification. When the introduction of the 5G base station's phase continuity band leads to a significant increase in the positioning residual, the sample is marked as a phase break. When the introduction of the 5G base station's phase continuity band does not lead to a significant increase in the positioning residual, and the trajectory changes in adjacent epochs remain continuous, the sample is marked as a phase continuation. Through the above method, the label definition has a clear correspondence with 5G phase continuity, BeiDou pseudorange direction constraints, and the final positioning residual, and has a clear physical meaning of positioning, rather than being arbitrarily labeled based on subjective experience.

[0029] The heterogeneous phase trusted solver is trained using supervised learning. The training samples are heterogeneous mutual attraction time series segments generated according to the aforementioned steps, and the sample labels include phase continuation labels and phase break labels. During training, the heterogeneous mutual attraction time series segments are input into the heterogeneous phase trusted solver. After processing by the mutual attraction encoding unit, phase consistency gating unit, and string bead evolution unit, the trusted label output unit outputs the continuous progressive state probability Pc and the interrupted state probability Pb.

[0030] During training, the total loss function is composed of classification cross-entropy loss and phase consistency constraint loss, expressed as: L = L1 + L2, where L is the total loss function, L1 is the classification cross-entropy loss, and L2 is the phase consistency constraint loss. The classification cross-entropy loss is expressed as: L1 = -[ylnPc + (1-y)lnPb], where y is the sample label; y is 1 in the phase continuation state and 0 in the phase break state. The phase consistency constraint loss is expressed as: L2 = yD + (1-y)max(0, 1.0 - D), where D is the chain-beam recursion feature distance between the current epoch and the adjacent previous epoch.

[0031] After training, when Pb is greater than Pc, a phase break marker is generated; when Pc is greater than or equal to Pb, a phase continuity marker is generated; for 5G base stations with phase continuity markers and a phase consistency gating coefficient greater than 0.65, a 5G phase trust marker is generated.

[0032] In this embodiment, a pseudorange phase bead recursion mechanism is implemented to recursively process the phase consistency gating features in adjacent solution epochs, specifically as follows: The phase consistency gating features of the same 5G base station in each solution epoch are obtained and arranged in the order of solution epochs to generate a single base station phase gating feature sequence. Set the phase-gated differential feature corresponding to the first epoch to zero vector. Starting from the second epoch, subtract the phase-consistency gated feature corresponding to the adjacent previous epoch from the phase-consistency gated feature corresponding to the current epoch to obtain the phase-gated differential feature corresponding to the current epoch. The phase consistency gating feature and the corresponding phase gating differential feature corresponding to each solution epoch are concatenated according to the channel dimension to obtain the temporal string unrolling feature; The temporal beam unpacking features are sequentially input into the first linear mapping layer, the LayerNorm layer, the GELU activation layer, and the second linear mapping layer to obtain the static beam evolution features of a single base station. The chain-beam evolution feature output by the same 5G base station in the previous adjacent solution epoch is taken as the historical state. The historical state is superimposed with the static beam evolution feature of the single base station corresponding to the current solution epoch to generate the chain-beam evolution feature of the current solution epoch. The historical state corresponding to the first solution epoch is initialized as a zero vector. According to the order of the solution epochs, the chain-like chord progression feature of the current solution epoch is used as the historical state of the next solution epoch and is rolled over until the last solution epoch generates the chain-like chord progression feature, which is used as the single base station chain progression feature of the corresponding 5G base station. In the specific implementation process, the pseudorange phase beamforming mechanism takes a single 5G base station as the processing object, and performs beamforming processing on the phase consistency gating features according to the order of each solution epoch. The number of channels for the phase consistency gating features is set to 32, and the number of channels for the phase gating differential features is set to 32. After being spliced ​​according to the channel dimension, a 64-channel temporal beamforming feature is formed. The first linear mapping layer is used to increase the dimensionality of the temporal beamforming feature. The input dimension is set to 64, and the output dimension is set to 128, so that the gating phase state of the current solution epoch and the change state of adjacent epochs enter the same high-dimensional representation space. The LayerNorm layer is used to normalize the 128-dimensional features to reduce the impact of the phase amplitude difference of different 5G base stations on the beamforming processing. The GELU activation layer is used to retain small phase evolution information and suppress instantaneous spike disturbances.

[0033] The second linear mapping layer compresses the 128-dimensional features after GELU activation back to 64 dimensions, obtaining the single-base station static beam evolution features. The historical state corresponding to the first participating epoch is set as a 64-dimensional zero vector. Subsequent epochs continuously use the current chain-beam evolution features as the historical state of the next epoch, thus preserving the preceding phase evolution information on the time axis. Through this processing, the single-base station chain-beam evolution features not only reflect the current phase state of the last participating epoch but also carry the continuous evolution traces of the preceding epochs, providing a stable feature basis for the trusted label output unit to distinguish between phase continuation labels and phase break labels.

[0034] In this embodiment, step five specifically includes: The continuous band of the base station phase corresponding to the 5G base station with 5G phase trust label is retained to obtain the trust phase continuous band; Delete the continuous phase band of the 5G base station corresponding to the phase break marker; The phase continuity band of the 5G base station with phase continuity mark but without 5G phase trust mark is taken as the phase continuity band to be corrected and directional consistency correction is performed. The number of positive and negative directions of the change in the continuous reconnection phase difference of each reliable phase continuous band within the same solution epoch is counted, and the direction with the larger number is determined as the reference change direction; The continuous back-loop phase difference value of the first participating epoch in the phase continuum to be corrected is taken as the corrected continuous back-loop phase difference value corresponding to the first participating epoch. Starting from the second participating epoch, the continuous back-loop phase difference value of the current solving epoch in the phase continuum to be corrected is subtracted from the continuous back-loop phase difference value of the adjacent previous solving epoch to obtain the phase change value to be corrected corresponding to the current solving epoch. When the positive and negative directions of the phase change to be corrected are consistent with the reference change direction, the continuous loop phase difference of the current solution epoch is taken as the corrected continuous loop phase difference of the current solution epoch. When the positive and negative directions of the phase change to be corrected are inconsistent with the reference change direction, the corrected continuous loop phase difference value corresponding to the adjacent previous solution epoch is used as the recursive reference. If the reference change direction is positive, the absolute value of the phase change to be corrected is added to the recursive reference. If the reference change direction is negative, the absolute value of the phase change to be corrected is subtracted from the recursive reference to obtain the corrected continuous loop phase difference value corresponding to the current solution epoch. By connecting the corrected phase difference values ​​in the order of the solution epochs, a corrected phase continuous band is obtained. The trusted phase continuous band and the modified phase continuous band are combined according to the 5G base station number to obtain the trusted 5G phase positioning constraint. In this invention, 5G phase trust markers are used to screen stable phase observations that can directly participate in subsequent calculations, phase break markers are used to exclude base station phase continuity bands that have experienced timing interruptions, and phase continuity markers are used to identify base station phase continuity bands that still have timing continuity but insufficient trust. For phase continuity bands to be corrected, they are not directly discarded, but are corrected using the dominant change direction presented by the trustworthy phase continuity bands within the same calculation epoch. This allows 5G base stations in a gray state to still participate in the construction of trustworthy 5G phase positioning constraints under controlled conditions, avoiding a reduction in the number of available 5G phase observations due to overly strict gating thresholds, and also preventing phase break data from entering the joint positioning calculation constraints.

[0035] Directional consistency correction is performed recursively. The first epoch in the calculation serves as the correction starting point, and subsequent epochs use the corrected phase difference corresponding to the previous epoch as the recursive reference. This ensures the correction process is continuously propagated chronologically. When the phase change to be corrected is consistent with the reference change direction, the original trend is retained. When they are inconsistent, only the change direction is adjusted while the change amplitude is retained. This ensures that the corrected phase continuity band follows the dominant direction of the reliable phase continuity band while retaining the phase change scale of the corresponding 5G base station itself. Through this process, reliable 5G phase positioning constraints can simultaneously include directly reliable phase observations and phase observations corrected for directional consistency, improving the continuity and availability of 5G phase constraints in complex environments.

[0036] In this embodiment, step six specifically includes: Extract the single-satellite spatial unit direction vector corresponding to the pseudorange contraction direction region from the pseudorange direction compression domain, and extract the corresponding single-satellite pseudorange difference as the BeiDou pseudorange residual, which together serve as the BeiDou pseudorange direction constraint. The continuous back-up phase difference corresponding to the trusted phase continuous band and the modified phase continuous band is extracted from the trusted 5G phase positioning constraint and used as the 5G phase continuity constraint. The 5G base station numbers participating in the solution are determined based on the 5G phase continuity constraints, and the 5G base station locations corresponding to the 5G base station numbers participating in the solution are obtained. The straight-line distances between the terminal's predicted coordinates and the corresponding 5G base station locations are calculated to obtain the base station predicted distances. Subtract the predicted distance of the corresponding base station from the continuous reconnection phase difference value of the 5G base station to obtain the continuous phase residual of 5G. A spatial geometric matrix is ​​established based on the single-star spatial unit direction vector in the pseudorange direction constraint of Beidou, without introducing the single-star spatial unit direction vector corresponding to the pseudorange exclusion direction region, thus limiting the direction update range of the terminal prediction coordinates. The spatial geometric matrix, BeiDou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and corresponding solution epoch are bound together to generate single epoch joint constraint terms; Arrange the joint constraint terms of each single epoch in the order of solution epochs to generate joint positioning solution constraints; In the specific implementation process, the spatial geometry matrix is ​​formed by arranging the single-satellite spatial unit direction vectors corresponding to the pseudorange contraction direction region. Each BeiDou satellite in the pseudorange contraction direction region corresponds to a row in the spatial geometry matrix, including the horizontal, vertical, and elevation components of the corresponding single-satellite spatial unit direction vector. The pseudorange difference between the single satellite and the corresponding single-satellite in that row of the spatial geometry matrix is ​​stored separately as the BeiDou pseudorange residual, and maintains a correspondence with the single-satellite spatial unit direction vector of the same BeiDou satellite. Thus, the spatial geometry matrix is ​​used to limit the spatial directions along which the terminal's predicted coordinates can be updated, and the BeiDou pseudorange residual is used to characterize the distance deviation in the corresponding direction. The single-satellite spatial unit direction vectors corresponding to the pseudorange exclusion direction region are not written into the spatial geometry matrix, and the corresponding single-satellite pseudorange difference is not included in the BeiDou pseudorange residual, thereby avoiding the influence of directionally dispersed or abnormal BeiDou pseudorange observations on the construction of joint positioning solution constraints. When calculating the predicted distance to the base station, the horizontal, vertical, and elevation coordinates of the terminal's predicted coordinates are subtracted from the horizontal, vertical, and elevation coordinates of the corresponding 5G base station location, respectively. The squares of these three coordinate differences are then summed and the square root is taken to obtain the straight-line distance to the corresponding 5G base station. Since the continuous reconnection phase difference and the predicted base station distance are on the same distance scale, the 5G phase continuous residual obtained by subtracting them can be directly used as a phase-side constraint. By binding the spatial geometric matrix, the BeiDou pseudorange residual, and the 5G phase continuous residual to the same solution epoch, the joint positioning solution constraint simultaneously preserves the directional stability of the BeiDou pseudorange and the continuous refinement capability of the 5G phase, improving the solvability of the terminal's predicted coordinate updates and its resistance to anomaly observations.

[0037] In this embodiment, step seven specifically includes: Read the single-epoch joint constraint terms of the corresponding solution epoch one by one from the joint positioning solution constraints, and extract the spatial geometric matrix, Beidou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and terminal predicted coordinates. Based on the single-star spatial unit direction vector in the spatial geometric matrix and the corresponding BeiDou pseudorange residual, the BeiDou coordinate correction is generated. The location of the corresponding 5G base station is obtained based on the corresponding 5G base station number, and the 5G coordinate correction amount is generated based on the spatial direction of the 5G base station location relative to the terminal's predicted coordinates and the 5G phase continuity residual. The BeiDou coordinate correction and the 5G coordinate correction are added together according to the horizontal coordinate, vertical coordinate and elevation coordinate to obtain the joint coordinate correction. The joint coordinate correction is superimposed on the terminal predicted coordinates according to the horizontal coordinate, vertical coordinate and elevation coordinate respectively, and the terminal predicted coordinates are updated to obtain the updated terminal predicted coordinates. The updated joint positioning solution constraints are then regenerated based on the updated terminal predicted coordinates. Calculate the coordinate change between the updated terminal predicted coordinates and the terminal predicted coordinates before the update; When the coordinate change exceeds the set coordinate change threshold, the terminal predicted coordinates continue to be updated based on the updated joint positioning solution constraints. When the coordinate change does not exceed the set coordinate change threshold, the updated terminal predicted coordinates will be determined as the joint positioning coordinates. In the specific implementation process, the spatial geometry matrix in the single-epoch joint constraint term only records the dimensionless single-satellite spatial unit direction vector, while the BeiDou pseudorange residual is a distance quantity. The two are kept separate in terms of direction and distance during computation. When generating the BeiDou coordinate correction, the lateral, longitudinal, and elevation components of each single-satellite spatial unit direction vector are multiplied by the corresponding BeiDou pseudorange residual to obtain the single-satellite lateral, longitudinal, and elevation correction components with distance dimensions. Then, all single-satellite lateral, longitudinal, and elevation correction components are summed and divided by the number of BeiDou satellites involved in the calculation to obtain the BeiDou coordinate correction.

[0038] When generating 5G coordinate corrections, the 5G phase continuous residual is the distance obtained by subtracting the predicted distance from the continuous reconnection phase difference. The lateral, longitudinal, and elevation components of the spatial unit direction vector for each base station are multiplied by the corresponding 5G phase continuous residual to obtain single-base station lateral, longitudinal, and elevation correction components with distance dimensions. Then, all single-base station lateral, longitudinal, and elevation correction components are summed and divided by the number of 5G base stations involved in the calculation to obtain the 5G coordinate correction. Since both BeiDou and 5G coordinate corrections are distance quantities, they can be added together according to the corresponding lateral, longitudinal, and elevation coordinates to obtain a joint coordinate correction, which is also a distance quantity.

[0039] Subsequently, the lateral component of the joint coordinate correction is superimposed onto the lateral coordinate of the terminal predicted coordinate, the longitudinal component of the joint coordinate correction is superimposed onto the longitudinal coordinate of the terminal predicted coordinate, and the elevation component of the joint coordinate correction is superimposed onto the elevation coordinate of the terminal predicted coordinate to obtain the updated terminal predicted coordinate. The coordinate change threshold is set to 0.02 meters, and the coordinate change is calculated according to the three-dimensional straight-line distance between the updated terminal predicted coordinate and the terminal predicted coordinate before the update. After each update of the terminal predicted coordinate, the updated joint positioning solution constraints are regenerated, so that the pseudorange direction compression domain, the trusted 5G phase positioning constraint, and the single epoch joint constraint terms are all updated synchronously with the coordinate change. This allows the BeiDou pseudorange direction constraint and the 5G phase continuity constraint to continuously converge collaboratively in the same iteration process, improving the stability of the joint positioning coordinate.

[0040] In this embodiment, step eight specifically includes: According to the order of solution epochs, obtain the joint positioning coordinates obtained by iterative solution in each solution epoch; Connect the joint positioning coordinates corresponding to adjacent solution epochs in chronological order to generate a joint positioning trajectory; Calculate the coordinate difference between two adjacent joint positioning coordinates to obtain the displacement of the trajectory segment; The joint positioning coordinates, joint positioning trajectory, trajectory segment displacement, and corresponding solution epochs are encapsulated and output as the joint positioning solution results.

[0041] Example 1: To verify the feasibility of this invention in practice, it was applied to a mobile terminal joint positioning test scenario in an area where elevated roads and densely built-up areas meet in a city. This area suffers from issues such as obstruction by tall buildings, reflections from roadside metal guardrails, uneven spatial distribution of 5G base stations, and interference from BeiDou pseudorange multipath propagation. Traditional single BeiDou pseudorange calculations are prone to lateral trajectory drift, and ordinary 5G-assisted positioning can produce trajectory jumps when phase measurements show integer discontinuities. Direct fusion methods tend to include abnormal pseudorange and broken phases in the calculation, leading to unstable joint positioning coordinates.

[0042] During the test, the mobile terminal continuously traveled along the road at a speed of approximately 30 km / h to 45 km / h, collecting BeiDou pseudorange, BeiDou satellite positions, 5G phase measurements, 5G base station positions, and terminal predicted coordinates within the same epoch. The epoch was set to 1 second, and 600 epochs were collected continuously. First, the pseudorange difference of a single satellite was calculated, and then classified according to the spatial neighborhood relationship between the BeiDou satellite and the terminal's predicted coordinates to generate a pseudorange directional compression domain. Subsequently, the phase difference of a single base station was continuously reconnected through the cumulative integer compensation to generate a set of phase continuity bands. For 5G base stations with weak spatial directional correspondence, they were marked as independent mutual referencing base stations through the base station mutual referencing index table to avoid forced binding. Phase continuity bands of base stations with phase break markers were not included in subsequent calculations. For phase continuity bands of base stations with phase continuity markers but without 5G phase reliability markers, directional consistency correction was performed to generate corrected phase continuity bands. Finally, joint positioning solution constraints were constructed based on BeiDou pseudorange direction constraints and 5G phase continuity constraints, and the terminal predicted coordinates were iteratively updated until the coordinate change did not exceed 0.02 meters, thus obtaining the joint positioning coordinates and joint positioning trajectory.

[0043] To verify the practical effect of the present invention, three comparison schemes were set up. Comparison scheme one is a single BeiDou pseudorange solution scheme, which directly uses the pseudorange of all visible BeiDou satellites for positioning without constructing a pseudorange direction compression domain; comparison scheme two is a direct combination scheme of BeiDou pseudorange and 5G phase, which directly constructs a joint residual from BeiDou pseudorange and 5G phase measurements without performing phase continuum set correction; comparison scheme three is a conventional filtering joint positioning scheme, which uses conventional Kalman filtering to fuse BeiDou pseudorange and 5G phase measurements, but does not set up a heterogeneous phase reliable solver and a pseudorange phase bead evolution mechanism. The comparison results are shown in Table 1.

[0044] Table 1. Comparison of Positioning Accuracy and Trajectory Stability of Different Positioning Solution Schemes

[0045] As shown in Table 1, the method of this invention outperforms the three comparative schemes in terms of average positioning error, 95% positioning error, maximum trajectory jump, trajectory continuity rate, and average number of iterations for coordinate convergence. The 95% positioning error refers to the error value located at the 95th percentile after arranging the positioning errors of all solution epochs from smallest to largest; it reflects the upper limit stability of the error for most positioning results in complex environments. Comparative Scheme 1 relies solely on BeiDou pseudorange calculation without filtering for abnormal pseudorange directions, resulting in an average positioning error of 2.86m, a 95% positioning error of 5.74m, and a maximum trajectory jump of 8.31m, indicating poor trajectory stability under obstructed and multipath environments. Comparative Scheme 2, although incorporating 5G phase measurement, fails to effectively and continuously reconnect the phase difference values ​​of a single base station or perform reliable filtering; therefore, the average positioning error remains at 1.94m, and the maximum trajectory jump is 5.26m. Comparative Scheme 3, employing conventional filtering and joint positioning, reduces the average positioning error to 1.32m and improves the trajectory continuity rate to 94.7%. However, due to the lack of a pseudorange direction compression domain, a heterogeneous phase reliable solver, and a pseudorange phase string bead evolution mechanism, it is still difficult to fully suppress the effects of phase breakage and abnormal pseudorange. The average positioning error of the method of this invention is reduced to 0.68m, and the 95% positioning error is reduced to 1.21m, indicating that the positioning error of 95% of the solution epochs is controlled within about 1.21m. The maximum trajectory jump is only 1.09m, the trajectory continuity rate reaches 98.6%, and the average number of coordinate convergence iterations is reduced to 3.6 times. This shows that the present invention can form more stable joint positioning solution constraints in complex environments and improve the accuracy, continuity and convergence efficiency of joint positioning coordinates.

[0046] This embodiment verifies that, even in complex environments such as BeiDou pseudorange obstruction, multipath interference, and integer discontinuities in 5G phase measurements, the present invention can effectively generate a pseudorange direction compression domain, a set of continuous phase bands, and reliable 5G phase positioning constraints, and stably update the terminal's predicted coordinates based on the joint positioning solution constraints. Experimental data shows that the method of the present invention can reduce the average positioning error and 95% positioning error, reduce the maximum trajectory jump, improve the trajectory continuity rate, and shorten the average number of iterations for coordinate convergence, indicating that the present invention can achieve a more stable, continuous, and reliable BeiDou-5G joint positioning solution.

[0047] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A joint positioning solution method based on BeiDou pseudorange and 5G phase measurement, characterized in that, Includes the following steps: Step 1: Calculate the pseudorange difference of a single satellite, and classify the pseudorange difference of a single satellite continuously according to the spatial neighborhood relationship of the Beidou satellite relative to the terminal's predicted coordinates, to generate a pseudorange direction compression domain; Step 2: Calculate the phase difference of a single base station, and continuously reconnect the phase difference of a single base station by accumulating integer compensation to generate a set of continuous phase bands; Step 3: Spatial cross-reference between the pseudorange direction compression domain and the phase continuous band set; based on the correspondence between the spatial direction of the 5G base station and the pseudorange direction compression domain, generate heterogeneous cross-reference timing segments. Step 4: Input the heterogeneous mutual reference timing segment into the heterogeneous phase trusted solver. The heterogeneous phase trusted solver includes a mutual reference coding unit, a phase consistency gating unit, a string beam evolution unit, and a trusted tag output unit. The string beam evolution unit sets up a pseudo-range phase string beam evolution mechanism to generate 5G phase trusted tags, phase break tags, and phase continuity tags. Step 5: Based on the 5G phase trust marker, phase break marker, and phase continuation marker, retain, delete, and correct the direction consistency of the phase continuous band set to obtain the trustworthy 5G phase positioning constraint; Step 6: Obtain the BeiDou pseudorange direction constraint based on the pseudorange direction compression domain, obtain the 5G phase continuity constraint based on the reliable 5G phase positioning constraint, and construct the joint positioning solution constraint; Step 7: Update the terminal predicted coordinates based on the joint positioning solution constraints, and iteratively generate updated joint positioning solution constraints until the coordinate change does not exceed the set coordinate change threshold, and obtain the joint positioning coordinates; Step 8: Connect the joint positioning coordinates according to the solution epoch order to obtain the joint positioning trajectory, and output the joint positioning solution result.

2. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step one specifically involves: Obtain the BeiDou pseudorange, BeiDou satellite position, and terminal predicted coordinates within the same solution epoch, and match the BeiDou pseudorange with the BeiDou satellite position according to the BeiDou satellite number; The straight-line distance between the terminal's predicted coordinates and the positions of each BeiDou satellite is calculated to obtain the single-satellite predicted geometric distance. The BeiDou pseudorange corresponding to each BeiDou satellite is then subtracted from the corresponding single-satellite predicted geometric distance to obtain the single-satellite pseudorange difference. Obtain the spatial line-of-sight vector of each BeiDou satellite position relative to the terminal's predicted coordinates, and normalize the spatial line-of-sight vector to obtain the single-satellite spatial unit direction vector; Map the spatial unit direction vector of each single satellite to the local celestial coordinate system with the terminal predicted coordinates as the origin, and calculate the spatial solid angle between the spatial unit direction vectors of any two Beidou satellites. BeiDou satellites whose spatial three-dimensional angle does not exceed the set neighborhood angle are classified into the same adjacent satellite group; The pseudorange difference values ​​of individual BeiDou satellites within the same adjacent satellite group are marked with positive and negative signs to obtain satellites with non-negative difference values ​​and satellites with negative difference values. When the number of BeiDou satellites in the same adjacent satellite group is greater than the set base threshold, and the number of non-negative difference satellites or negative difference satellites in the group is greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is determined as a directional continuous adjacent group. When the number of BeiDou satellites in an adjacent satellite group is not greater than the set base threshold, or when the number of non-negative difference satellites and the number of negative difference satellites in the group are both not greater than the set percentage of the number of BeiDou satellites in the adjacent satellite group, the adjacent satellite group is defined as a directionally dispersed adjacent group. The single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of continuous adjacency group are classified into the pseudorange contraction direction region, and the single-star spatial unit direction vector and the single-star pseudorange difference corresponding to the direction of dispersed adjacency group are classified into the pseudorange repulsion direction region. The pseudorange contraction direction region and the pseudorange repulsion direction region are bound together according to the same solution epoch to generate a pseudorange direction compression domain.

3. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step two specifically involves: Obtain 5G phase measurement values, 5G base station locations, 5G signal wavelengths, and terminal predicted coordinates within the same solution epoch; Multiply the 5G phase measurement value by the corresponding 5G signal wavelength to obtain the initial phase distance of a single base station; Calculate the straight-line distance between the terminal's predicted coordinates and the location of each 5G base station to obtain the predicted geometric distance of a single base station, and subtract the predicted geometric distance of the corresponding single base station from the initial phase distance of each 5G base station to obtain the phase difference of the single base station. Arrange the single base station phase difference values ​​corresponding to the same 5G base station in the order of the solution epochs, set the cumulative integer compensation amount of the same 5G base station in the first solution epoch to zero, and take the single base station phase difference value in the first solution epoch as the continuous reconnection phase difference value in the first solution epoch. Starting from the second epoch involved in the solution, the difference between the single-base station phase difference of the current solution epoch and the single-base station phase difference of the adjacent previous solution epoch is calculated to obtain the phase evolution difference between epochs. When the phase evolution difference between epochs is greater than the set positive cycle jump threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is increased by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is less than the set reverse cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is reduced by one 5G signal wavelength to obtain the cumulative integer compensation amount corresponding to the current solution epoch. When the phase evolution difference between epochs is between the set reverse cycle slip threshold and the set forward cycle slip threshold, the cumulative integer compensation amount corresponding to the adjacent previous solution epoch is used as the cumulative integer compensation amount corresponding to the current solution epoch. The single base station phase difference value of the current solution epoch is deducted and corrected based on the cumulative integer compensation amount corresponding to the current solution epoch to obtain the continuous reconnection phase difference value. The continuous phase difference values ​​of the same 5G base station in the continuous solution epoch are connected in the order of solution epoch to obtain the continuous phase band of the base station, and the continuous phase bands of each base station are summarized to generate a continuous phase band set.

4. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step three specifically involves: Subtract the terminal's predicted coordinates from the 5G base station location within the same solution epoch to obtain the base station spatial line-of-sight vector. Then, normalize the base station spatial line-of-sight vector to generate the 5G base station spatial unit direction vector. Calculate the spatial solid angle between the spatial unit direction vector of the 5G base station and the spatial unit direction vector of each single star in the pseudorange direction compression domain to obtain the set of base station star direction angles. Select the smallest spatial three-dimensional angle from the set of base station star orientation angles, and determine the single-star spatial unit direction vector corresponding to the smallest spatial three-dimensional angle as the mutual attraction BeiDou direction of the 5G base station, and determine the smallest spatial three-dimensional angle as the mutual attraction adaptation angle. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange contraction direction area in the pseudorange direction compression domain, the 5G base station is marked as a contraction mutual attraction base station. When the mutual attraction adaptation angle does not exceed the set spatial mutual attraction angle threshold, and the mutual attraction BeiDou direction belongs to the pseudorange exclusion direction region in the pseudorange direction compression domain, the 5G base station will be marked as a mutually exclusive base station. When the mutual referencing and adaptation angle exceeds the set spatial mutual referencing angle threshold, the 5G base station will be marked as an independent mutual referencing base station. According to the 5G base station number, the marking results of shrinking mutual referencing base stations, excluding mutual referencing base stations, and independent mutual referencing base stations are written into the base station mutual referencing index table; Extract the phase continuous bands of each base station corresponding to the base station mutual reference index table from the phase continuous band set, and calculate the change in the continuous reconnection phase difference between the current solution epoch and the adjacent previous solution epoch in each base station phase continuous band. The base station mutual reference index table and the change in phase difference of continuous reconnection are spatiotemporally bound according to the solution epoch to generate heterogeneous mutual reference time sequence segments.

5. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step four specifically involves: The heterogeneous mutual reference time sequence segment is input into the mutual reference encoding unit in the heterogeneous phase reliable solver. The base station mutual reference index table and the change in continuous reconnection phase difference in the heterogeneous mutual reference time sequence segment are vectorized and arranged to obtain the mutual reference input vector. The mutual reference input vector is input into a one-dimensional convolutional layer for encoding, and then non-linearly activated by the GELU activation function to obtain the mutual reference encoded features; The mutual referencing coding features are input into the phase consistency gating unit, and mutual referencing type coding is generated for each 5G base station based on the marking results of shrinking mutual referencing base stations, exclusion mutual referencing base stations and independent mutual referencing base stations in the base station mutual referencing index table. The mutual reference type code and the corresponding mutual reference code feature of the 5G base station are concatenated according to the channel dimension to obtain the mutual reference type concatenation feature; The concatenated features of mutual reference types are input into the gated mapping layer and phase consistency gating coefficients are generated by the Sigmoid activation function. The phase consistency gating coefficients are multiplied element-wise with the cross-reference coding features to obtain the phase consistency gating features; The phase consistency gating features are input into the string beam evolution unit, and the pseudorange phase string beam evolution mechanism is executed to perform evolution processing on the phase consistency gating features in adjacent solution epochs to obtain the single base station chain evolution features of each 5G base station. The single-base station chain evolution feature of each 5G base station is input into the trusted label output unit. Through the linear mapping layer, the evolution state mapping vector is obtained. The evolution state mapping vector is then input into the Softmax function to obtain the continuous progressive state probability and interruption state probability of the corresponding 5G base station. When the probability of an interrupted state is greater than the probability of a continuous progressive state, a phase break marker is generated for the corresponding 5G base station; when the probability of a continuous progressive state is greater than or equal to the probability of an interrupted state, a phase continuation marker is generated for the corresponding 5G base station. 5G base stations with phase continuity markings and corresponding phase consistency gating coefficients greater than the set gating threshold are marked as 5G phase trust markings.

6. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 5, characterized in that, The pseudorange phase bead recursion mechanism performs recursive processing on the phase consistency gating features in adjacent solution epochs, specifically as follows: The phase consistency gating features of the same 5G base station in each solution epoch are obtained and arranged in the order of solution epochs to generate a single base station phase gating feature sequence. Set the phase-gated differential feature corresponding to the first epoch to zero vector. Starting from the second epoch, subtract the phase-consistency gated feature corresponding to the adjacent previous epoch from the phase-consistency gated feature corresponding to the current epoch to obtain the phase-gated differential feature corresponding to the current epoch. The phase consistency gating feature and the corresponding phase gating differential feature corresponding to each solution epoch are concatenated according to the channel dimension to obtain the temporal string unrolling feature; The temporal beam unpacking features are sequentially input into the first linear mapping layer, the LayerNorm layer, the GELU activation layer, and the second linear mapping layer to obtain the static beam evolution features of a single base station. The chain-beam evolution feature output by the same 5G base station in the previous adjacent solution epoch is taken as the historical state. The historical state is superimposed with the static beam evolution feature of the single base station corresponding to the current solution epoch to generate the chain-beam evolution feature of the current solution epoch. The historical state corresponding to the first solution epoch is initialized as a zero vector. According to the order of the solution epochs, the chain-like chord progression feature of the current solution epoch is used as the historical state of the next solution epoch and is rolled over until the last solution epoch generates the chain-like chord progression feature, which is then used as the single-base station chain progression feature of the corresponding 5G base station.

7. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step five specifically involves: The continuous band of the base station phase corresponding to the 5G base station with 5G phase trust label is retained to obtain the trust phase continuous band; Delete the continuous phase band of the 5G base station corresponding to the phase break marker; The phase continuity band of the 5G base station with phase continuity mark but no 5G phase trust mark is taken as the phase continuity band to be corrected and directional consistency correction is performed. The number of positive and negative directions of the change in the continuous reconnection phase difference of each reliable phase continuous band within the same solution epoch is counted, and the direction with the larger number is determined as the reference change direction; The continuous back-loop phase difference value of the first participating epoch in the phase continuum to be corrected is taken as the corrected continuous back-loop phase difference value corresponding to the first participating epoch. Starting from the second participating epoch, the continuous back-loop phase difference value of the current solving epoch in the phase continuum to be corrected is subtracted from the continuous back-loop phase difference value of the adjacent previous solving epoch to obtain the phase change value to be corrected corresponding to the current solving epoch. When the positive and negative directions of the phase change to be corrected are consistent with the reference change direction, the continuous loop phase difference of the current solution epoch is taken as the corrected continuous loop phase difference of the current solution epoch. When the positive and negative directions of the phase change to be corrected are inconsistent with the reference change direction, the corrected continuous reconnection phase difference value corresponding to the adjacent previous solution epoch is used as the recursive reference. If the reference change direction is positive, the absolute value of the phase change to be corrected is added to the recursive reference. If the reference change direction is negative, the absolute value of the phase change to be corrected is subtracted from the recursive reference to obtain the corrected continuous reconnection phase difference value corresponding to the current solution epoch. By connecting the corrected phase difference values ​​in the order of the solution epochs, a corrected phase continuous band is obtained. The trusted phase continuous band and the modified phase continuous band are combined according to the 5G base station number to obtain the trusted 5G phase positioning constraint.

8. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step six specifically involves: Extract the single-satellite spatial unit direction vector corresponding to the pseudorange contraction direction region from the pseudorange direction compression domain, and extract the corresponding single-satellite pseudorange difference as the BeiDou pseudorange residual, which together serve as the BeiDou pseudorange direction constraint. The continuous back-up phase difference corresponding to the trusted phase continuous band and the modified phase continuous band is extracted from the trusted 5G phase positioning constraint and used as the 5G phase continuity constraint. The 5G base station numbers participating in the solution are determined based on the 5G phase continuity constraints, and the 5G base station locations corresponding to the 5G base station numbers participating in the solution are obtained. The straight-line distances between the terminal's predicted coordinates and the corresponding 5G base station locations are calculated to obtain the base station predicted distances. Subtract the predicted distance of the corresponding base station from the continuous reconnection phase difference value of the 5G base station to obtain the continuous phase residual of 5G. A spatial geometric matrix is ​​established based on the single-star spatial unit direction vector in the pseudorange direction constraint of Beidou, without introducing the single-star spatial unit direction vector corresponding to the pseudorange exclusion direction region, thus limiting the direction update range of the terminal prediction coordinates. The spatial geometric matrix, BeiDou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and corresponding solution epoch are bound together to generate single epoch joint constraint terms; Arrange the joint constraint terms of each single epoch in the order of solution epochs to generate joint positioning solution constraints.

9. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step seven specifically involves: Read the single-epoch joint constraint terms of the corresponding solution epoch one by one from the joint positioning solution constraints, and extract the spatial geometric matrix, Beidou pseudorange residual, 5G phase continuous residual, corresponding 5G base station number and terminal predicted coordinates. Based on the single-star spatial unit direction vector in the spatial geometric matrix and the corresponding BeiDou pseudorange residual, the BeiDou coordinate correction is generated. The location of the corresponding 5G base station is obtained based on the corresponding 5G base station number, and the 5G coordinate correction amount is generated based on the spatial direction of the 5G base station location relative to the terminal's predicted coordinates and the 5G phase continuity residual. The BeiDou coordinate correction and the 5G coordinate correction are added together according to the horizontal coordinate, vertical coordinate and elevation coordinate to obtain the joint coordinate correction. The joint coordinate correction is superimposed on the terminal predicted coordinates according to the horizontal coordinate, vertical coordinate and elevation coordinate respectively, and the terminal predicted coordinates are updated to obtain the updated terminal predicted coordinates. The updated joint positioning solution constraints are then regenerated based on the updated terminal predicted coordinates. Calculate the coordinate change between the updated terminal predicted coordinates and the terminal predicted coordinates before the update; When the coordinate change exceeds the set coordinate change threshold, the terminal predicted coordinates continue to be updated based on the updated joint positioning solution constraints. When the coordinate change does not exceed the set coordinate change threshold, the updated terminal predicted coordinates will be determined as the joint positioning coordinates.

10. The joint positioning solution method based on BeiDou pseudorange and 5G phase measurement according to claim 1, characterized in that, Step eight specifically involves: According to the order of solution epochs, obtain the joint positioning coordinates obtained by iterative solution in each solution epoch; Connect the joint positioning coordinates corresponding to adjacent solution epochs in chronological order to generate a joint positioning trajectory; Calculate the coordinate difference between two adjacent joint positioning coordinates to obtain the displacement of the trajectory segment; The joint positioning coordinates, joint positioning trajectory, trajectory segment displacement, and corresponding solution epochs are encapsulated and output as joint positioning solution results.

Images