Positioning method and device, storage medium and navigation system

Abstract

The invention relates to a positioning method and device, a storage medium and a navigation system. The positioning method comprises the following steps: acquiring low earth orbit satellite information and positioning system observation information corresponding to a plurality of low earth orbit satellites at the current moment; performing topology modeling processing based on the low-orbit satellite information and a preset dynamic prediction model to obtain a satellite feature prediction matrix corresponding to the next moment of the current moment; performing multi-source data fusion processing based on the satellite feature prediction matrix, the low-orbit satellite information, the positioning system observation information and a preset Kalman filter to obtain a floating-point number result corresponding to the current position; and based on the floating-point number result and a preset ambiguity fixed model, carrying out positioning calculation processing to obtain a target positioning result. According to the invention, more stable and comprehensive low-orbit satellite information can be obtained, topology modeling is carried out through the preset dynamic prediction model, frequent switching of inter-satellite links caused by high-speed movement of low-orbit satellites is overcome, and the stability of the positioning model during topology change is ensured.

Description

Technical Field

[0001] This disclosure relates to the field of navigation technology, and in particular to positioning methods, devices, storage media and navigation systems. Background Technology

[0002] Low Earth Orbit (LEO) satellites, due to their rapid geometric configuration changes and strong signal power, have become a key means to improve the positioning accuracy and convergence speed of Global Navigation Satellite System (GNSS). Current mainstream technical solutions include joint positioning models: First, by constructing joint observation equations between LEO and medium- and high-orbit satellites, the positioning capability in polar regions can be improved. Second, by utilizing LEO satellites to simultaneously provide communication and navigation services, the signal coverage problem in remote areas can be solved, such as vehicle-mounted terminals using LEO to forward GNSS augmentation information to achieve high-precision positioning. Third, by using edge artificial intelligence modules to achieve onboard real-time object detection and data compression, downlink transmission delay can be reduced. However, existing technologies still have the following problems: (1) the quality of observation data is unstable, and there is redundancy and missing data in onboard GNSS data and difficulty in capturing high-dynamic signals; (2) the high-speed movement of LEO satellites relative to the ground results in short visible arcs and frequent link switching, leading to positioning interruptions. Summary of the Invention

[0003] To address at least one of the aforementioned technical problems, this disclosure provides a positioning method, apparatus, storage medium, and navigation system.

[0004] According to one aspect of this disclosure, a positioning method is provided for use in a navigation system, the navigation system comprising multiple low-Earth orbit (LEO) satellites, wherein the LEO satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration, comprising: Acquire multi-source observation data corresponding to the current moment, wherein the multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites; Based on the low-Earth orbit satellite information and the preset dynamic prediction model, topology modeling is performed to obtain the satellite feature prediction matrix corresponding to the next time step of the current time step. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on the low-Earth orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on the low-Earth orbit satellite information. Based on the satellite feature prediction matrix, the low-orbit satellite information, the positioning system observation information, and the preset Kalman filter, multi-source data fusion processing is performed to obtain the floating-point result corresponding to the current position; Based on the floating-point result, the low-orbit satellite information, and the preset ambiguity fixed model, the positioning calculation is performed to obtain the target positioning result.

[0005] In some possible implementations, the step of performing topology modeling based on the low-Earth orbit satellite information and a preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step at the current time step includes: Based on the low-Earth orbit satellite information, a satellite feature matrix corresponding to the current time is constructed. The satellite feature matrix is ​​used to characterize the state features corresponding to the multiple low-Earth orbit satellites. A satellite adjacency matrix is ​​constructed based on the low-Earth orbit satellite information. The satellite adjacency matrix is ​​used to characterize the spatial dependencies between the multiple low-Earth orbit satellites. A weighted aggregation process is performed based on the satellite adjacency matrix and the satellite feature matrix to obtain an initial spatial fusion matrix; Based on the initial spatial fusion matrix and the spatial feature transformation matrix, feature transformation processing is performed to obtain the spatial feature matrix.

[0006] In some possible implementations, the method further includes: The initial time fusion matrix is ​​obtained by weighted aggregation of the temporal convolution kernel parameter matrix of the preset dynamic prediction model and the satellite feature matrix. Based on the initial time fusion matrix and the time feature transformation matrix, feature transformation processing is performed to obtain the time feature matrix.

[0007] In some possible implementations, the low-Earth orbit satellite information includes Doppler frequency shift, and the multi-source data fusion processing based on the satellite feature prediction matrix, the low-Earth orbit satellite information, the positioning system observation information, and a preset Kalman filter to obtain the floating-point result corresponding to the current position includes: Based on the satellite feature prediction matrix and the Doppler frequency shift, update the noise matrix corresponding to the observation equation and the constraint matrix corresponding to the observation equation in the preset Kalman filter to obtain the updated observation equation and the updated constraint matrix; The preset Kalman filter is updated based on the updated observation equation and the updated constraint matrix to obtain the updated preset Kalman filter; The floating-point result is obtained by performing multi-source data fusion processing based on the low-orbit satellite information, the positioning system observation information, and the updated preset Kalman filter.

[0008] In some possible implementations, the positioning calculation based on the floating-point result, the low-Earth orbit satellite information, and a preset ambiguity fixed model to obtain the target positioning result includes: Based on the low-Earth orbit satellite information, the solution state characteristic parameters are determined, and the solution state characteristic parameters are used to characterize the signal quality characteristics corresponding to multiple low-Earth orbit satellites. The test threshold is determined based on the floating-point result and the solution state characteristic parameters; Based on the floating-point results, an evaluation process is performed to obtain a candidate set of integer solutions; The target localization result is determined based on the test threshold and the integer solution candidate set.

[0009] In some possible implementations, the floating-point result includes an ambiguity parameter and a residual sequence corresponding to the ambiguity parameter, and the step of determining the test threshold based on the floating-point result and the solution state feature parameter includes: Local features are extracted based on the residual sequence corresponding to the ambiguity parameters to obtain local features. Based on the ambiguity parameters and the solution state characteristic parameters, time series characteristic analysis is performed to obtain time series characteristics; The target fusion feature is obtained by fusing the local features and the temporal features. The target fusion features are subjected to threshold adaptive mapping to obtain the test threshold.

[0010] In some possible implementations, the integer candidate set includes optimal candidate solutions and suboptimal candidate solutions, and determining the target localization result based on the test threshold and the integer candidate solution set includes: The target ratio is determined based on the optimal candidate solution and the second-best candidate solution; A fixed result is determined based on the test threshold and the target ratio; If the fixation result is successful, the position is solved based on the optimal candidate solution and the updated observation equation to obtain the target positioning result.

[0011] According to a second aspect of this disclosure, a positioning device is provided for use in a navigation system, the navigation system comprising a plurality of low-Earth orbit (LEO) satellites, wherein the LEO satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration, the device comprising: The information acquisition module is used to acquire multi-source observation data corresponding to the current moment. The multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites. The topology modeling module is used to perform topology modeling processing based on the low-Earth orbit satellite information and a preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step of the current time step. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on the low-Earth orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on the low-Earth orbit satellite information. The floating-point number calculation module is used to perform multi-source data fusion processing based on the satellite feature prediction matrix, the low-orbit satellite information, the positioning system observation information, and the preset Kalman filter to obtain the floating-point number result corresponding to the current position; The positioning result determination module is used to perform positioning calculation processing based on the floating-point number result, the low-orbit satellite information and the preset ambiguity fixed model to obtain the target positioning result.

[0012] According to a third aspect of this disclosure, a navigation system is provided, the navigation system comprising a plurality of low-Earth orbit satellites and a positioning device as described in the second aspect, wherein the plurality of low-Earth orbit satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration.

[0013] According to a fourth aspect of this disclosure, an electronic device is provided, including at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the at least one processor implements the positioning method as described in any one of the first aspects by executing the instructions stored in the memory.

[0014] According to a fifth aspect of this disclosure, a computer-readable storage medium is provided that stores at least one instruction or at least one program, the at least one instruction or at least one program being loaded and executed by a processor to implement the positioning method as described in any of the first aspects.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure.

[0016] Implementing this disclosure will have the following beneficial effects: The navigation system comprises multiple low-Earth orbit (LEO) satellites, deployed in a hybrid polar-orbiting and non-polar-orbiting configuration. Building upon the existing polar-orbiting configuration, a polar-orbiting plus inclined orbit deployment method is employed to enhance mid-latitude and longitude coverage density. Multi-source observation data corresponding to the current moment is acquired, including LEO satellite information and positioning system observation information. By adding inclined orbit satellites, the spatial geometric distribution diversity of satellites is physically enhanced, providing richer and more stable LEO satellite information for downstream computation. Based on the LEO satellite information and a pre-defined dynamic prediction model, topology modeling is performed to obtain the satellite feature prediction matrix for the next moment. This matrix is ​​a fusion of spatial and temporal feature matrices. The spatial feature matrix is ​​obtained by extracting spatial features based on LEO satellite information, while the temporal feature matrix is ​​obtained by extracting temporal features based on LEO satellite information. By capturing spatial dependencies through the spatial feature matrix and learning the temporal patterns of satellite motion through the temporal feature matrix, the satellite feature prediction matrix incorporates a high-level representation of satellite state with spatiotemporal characteristics. Multi-source data fusion processing is performed based on satellite feature prediction matrices, low-Earth orbit satellite information, positioning system observation information, and a preset Kalman filter to obtain the floating-point result corresponding to the current position. This multi-source data fusion processing addresses the issue of unstable observation data quality and reduces positioning errors. Positioning calculation processing is then performed based on the floating-point result, low-Earth orbit satellite information, and a preset ambiguity fixing model to obtain the target positioning result. The preset ambiguity fixing model solves the problem of low ambiguity fixing efficiency, improving both positioning efficiency and accuracy.

[0017] Other features and aspects of this disclosure will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of this application, the accompanying drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 A flowchart illustrating a positioning method according to an embodiment of the present disclosure is shown; Figure 2 A flowchart illustrating a method for determining a spatial feature matrix according to an embodiment of the present disclosure is shown. Figure 3 A flowchart illustrating a method for determining a time feature matrix according to an embodiment of the present disclosure is shown. Figure 4A flowchart illustrating a method for determining floating-point results according to an embodiment of the present disclosure is shown. Figure 5 A flowchart illustrating a method for determining target location results according to an embodiment of the present disclosure is shown. Figure 6 A flowchart illustrating a method for determining an inspection threshold according to an embodiment of the present disclosure is shown. Figure 7 A schematic flowchart of a position determination method according to an embodiment of the present disclosure is shown; Figure 8 This diagram shows a structural schematic of a positioning device according to an embodiment of the present disclosure; Figure 9 A block diagram of an electronic device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0020] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0022] Various exemplary embodiments, features, and aspects of this disclosure will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0023] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments.

[0024] In this document, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.

[0025] Furthermore, to better illustrate this disclosure, numerous specific details are set forth in the following detailed description. Those skilled in the art will understand that this disclosure can be practiced without certain specific details. In some instances, methods, means, components, and circuits well known to those skilled in the art have not been described in detail in order to highlight the main points of this disclosure.

[0026] Figure 1 This diagram illustrates a flowchart of a positioning method according to an embodiment of the present disclosure, applied to a navigation system. The navigation system includes multiple low-Earth orbit (LEO) satellites, and the LEO satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration, such as... Figure 1 As shown, it includes: S101. Obtain the multi-source observation data corresponding to the current moment. The multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites. Polar orbit deployment involves an orbital tilt angle of approximately 90 degrees, while non-polar orbit deployment involves a tilted deployment method.

[0027] The navigation system adds a preset number of low-Earth orbit satellites with a preset tilt angle to the existing polar orbit low-Earth orbit satellites. The preset number can be 12 and the preset angle can be 45 degrees.

[0028] Multi-source observation data includes, but is not limited to, low-Earth orbit (LEO) satellite information, positioning system observation information, laser ranging point cloud data, and inertial measurement unit (IMU) data. LEO satellite information includes, but is not limited to, LEO satellite observation data and ephemeris data. LEO satellite observation data includes, but is not limited to, pseudorange, carrier phase, and Doppler shift. Ephemeris data includes, but is not limited to, orbital parameters. Positioning system observation information includes, but is not limited to, Global Positioning System (GPS) observation data and BeiDou Navigation Satellite System (BDS) observation data.

[0029] S102. Based on low-orbit satellite information and a preset dynamic prediction model, topology modeling is performed to obtain the satellite feature prediction matrix corresponding to the next moment of the current moment. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on low-orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on low-orbit satellite information. The preset dynamic prediction model is trained based on Dynamic Graph Neural Networks (DGNN). Based on low-Earth orbit (LEO) satellite observation data, the satellite feature matrix corresponding to the LEO satellite at the current moment is determined. Each row of the satellite feature matrix corresponds to the feature vector of a LEO satellite. The feature vector includes, but is not limited to, position feature elements, velocity feature elements, and signal strength elements.

[0030] In some embodiments, the core formula of the preset dynamic prediction model is as follows:

[0031] in, This is the fused satellite feature prediction matrix for the next time step. This is the feature matrix of N satellites at the current time, with dimensions N×F, where N is the number of satellites and F is the original feature dimension. , This is the dynamic adjacency matrix corresponding to the current moment, used to describe the spatial relationships between satellites. The larger the value, the stronger the spatial correlation between satellite i and satellite j. By aggregating the features of neighboring satellites through matrix multiplication, such as weighted averaging, spatial dependencies can be captured. , The spatial feature transformation matrix maps the aggregated features from the original F-dimensional space to a new F′-dimensional space through a linear transformation. Learning more effective feature representations is essentially about training a parameter matrix, which is then optimized during the training of the prediction model. , The temporal convolution kernel is defined by the following formula:

[0032] This is a time-feature variation matrix. This is the time-feature transformation matrix.

[0033] S103. Based on the satellite feature prediction matrix, low-orbit satellite information, positioning system observation information and preset Kalman filter, multi-source data fusion processing is performed to obtain the floating-point result corresponding to the current position; The preset Kalman filter is trained based on a Kalman filter with introduced Doppler constraints. The noise matrix and its corresponding control constraints in the introduced Doppler constraints of the preset Kalman filter are updated based on the satellite feature prediction matrix, resulting in an updated preset Kalman filter. Multi-source data fusion processing is then performed based on low-Earth orbit satellite information, positioning system observation information, and the updated preset Kalman filter to obtain the floating-point result corresponding to the current position. The floating-point result is used to characterize the preliminary fusion result of the multi-source data. The floating-point result includes, but is not limited to, the three-dimensional coordinates, velocity, clock error, covariance matrix, floating-point unambiguity parameter N_ambiguity, and its corresponding residual sequence N_residual of the receiver to be positioned. The covariance matrix is ​​used to characterize the uncertainty.

[0034] In some embodiments, similarity or distance analyses are performed on the satellite feature prediction matrix for the next time step to predict link status information. This predicted status information can indicate the existence of a stable, visible link between the low-Earth orbit satellite and the receiver to be located at the current time, as well as the expected link quality, such as signal-to-noise ratio estimation. Kalman filter parameters are adjusted based on this predicted status information to achieve adaptive adjustment of the Kalman filter. The observation noise matrix R is also adjusted based on the predicted status information. For Doppler observations of low-Earth orbit satellites with good predicted link status and high stability, the system assigns them higher weights, i.e., a smaller first variance value is allocated to them in the noise matrix R, indicating greater trust in the observation. Conversely, for observations of satellites with unstable or impending link failures, a larger second variance value is assigned to the R matrix to reduce their influence in filtering; the second variance value is greater than the first variance value. The strength of the control constraint term B is also adjusted based on the predicted status information. The constraint matrix B is not fixed; its strength, or the tightness of the constraint, can be dynamically adjusted according to the link quality in the predicted status information. When the predicted link quality is excellent, the strength of the Doppler constraint can be enhanced, which is equivalent to increasing the gain of B, making its correction effect on the state estimation more significant. After the above-mentioned adaptive parameter adjustment, the Kalman filter performs the prediction and update steps.

[0035] S104. Based on the floating-point results, low-orbit satellite information, and a preset ambiguity fixed model, the positioning solution is performed to obtain the target positioning result.

[0036] The preset ambiguity fixing model can be obtained by training a dual-channel model based on a convolutional neural network (CNN) and a long short-term memory (LSTM) network. The solution state feature parameters are determined based on low-Earth orbit satellite information. The solution state feature parameters include, but are not limited to, the position dilution of precision (PDOP) and signal-to-noise ratio (SNR) based on time sequence arrangement.

[0037] In some embodiments, convolution operations are performed based on floating-point results and CNN channels in a preset ambiguity fixed model to obtain local features (Feature_CNN). The temporally ordered PDOP values ​​and floating-point unambiguity parameters (N_ambiguity) are analyzed and processed through LSTM channels in the preset ambiguity fixed model to obtain temporal features. A test threshold is determined based on local features and temporal features. The target localization result is determined based on the test threshold and floating-point results.

[0038] The aforementioned technical solution, based on the existing polar orbit, adopts a polar orbit plus inclined orbit deployment method to improve the coverage density in mid-latitude and longitude regions. By adding inclined orbit satellites, it physically enhances the diversity of satellite spatial geometric distribution, providing richer and more stable low-Earth orbit satellite information for downstream computing. Spatial dependencies are captured through a spatial feature matrix, and the temporal feature matrix learns the temporal patterns of satellite motion, enabling the satellite feature prediction matrix to incorporate a high-level representation of satellite state with spatiotemporal features. Multi-source data fusion processing addresses the problem of unstable observation data quality, reducing positioning errors. A pre-set ambiguity fixing model solves the problem of low ambiguity fixing efficiency, improving positioning efficiency and accuracy.

[0039] Please see Figure 2 In some embodiments, topology modeling is performed based on low-Earth orbit satellite information and a preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step, including: S1021. Construct a satellite feature matrix corresponding to the current moment based on low-orbit satellite information. The satellite feature matrix is ​​used to characterize the state features of multiple low-orbit satellites. S1022. Construct a satellite adjacency matrix based on low-Earth orbit satellite information. The satellite adjacency matrix is ​​used to characterize the spatial dependency relationship between multiple low-Earth orbit satellites. S1023. Based on the satellite adjacency matrix and satellite feature matrix, a weighted aggregation process is performed to obtain the initial spatial fusion matrix; S1024. Based on the initial spatial fusion matrix and the spatial feature transformation matrix, feature transformation processing is performed to obtain the spatial feature matrix.

[0040] Low-Earth orbit (LEO) satellite observation data and ephemeris data from different sources are converted into a structured numerical form to obtain a satellite feature matrix. Each row of the satellite feature matrix corresponds to a feature vector of a LEO satellite, and the feature vector includes, but is not limited to, position feature elements, velocity feature elements, and signal strength elements. Inter-satellite ranging is performed based on the LEO satellite observation data and ephemeris data to construct an adjacency matrix. If the distance between satellite i and satellite j is less than a preset distance and there are no other obstructing signals, a usable link is determined to exist between them. Physical quantities such as signal-to-noise ratio and the reciprocal of distance can be used as weights. The larger the value, the stronger the spatial correlation between satellite i and satellite j. Features of neighboring satellites are aggregated through matrix multiplication, and the dynamic adjacency matrix Φ is used to weight and aggregate satellite features, ensuring that each satellite's features include information about its neighbors. The spatial feature transformation matrix is ​​a pre-trained parameter matrix in the model. Features aggregated through the spatial feature transformation matrix undergo dimensional transformation and optimization to extract higher-level abstract features, extracting spatial motion patterns from the original position / velocity features.

[0041] In some embodiments, the core formula of the preset dynamic prediction model is as follows:

[0042] in, , This is the spatial feature matrix after spatial feature extraction. The time feature matrix is ​​the time feature matrix after time feature extraction.

[0043] =

[0044] in, This is the satellite feature matrix corresponding to N satellites at the current time. , This is the satellite adjacency matrix corresponding to the current time. This is the spatial feature transformation matrix.

[0045] The above technical solution integrates spatial correlation with the current satellite state characteristics and aggregates information from neighboring satellites through an adjacency matrix, thereby better updating the satellite's own state characteristics, capturing spatial dependencies, predicting dynamic changes in network topology, and providing more accurate routing decision information.

[0046] Please see Figure 3 In some embodiments, the method further includes: S201. The initial time fusion matrix is ​​obtained by weighted aggregation of the temporal convolution kernel parameter matrix and satellite feature matrix based on the preset dynamic prediction model. S202. Based on the initial time fusion matrix and the time feature transformation matrix, feature transformation processing is performed to obtain the time feature matrix.

[0047] The temporal convolution kernel parameter matrix is ​​a trainable parameter matrix initialized before model training and whose optimal values ​​are learned from data during the training of the preset dynamic prediction model. The temporal feature transformation matrix is ​​also a trainable parameter matrix whose optimal values ​​are learned from data during the training of the preset dynamic prediction model.

[0048] In some embodiments, the formula for the time feature matrix is ​​as follows: =

[0049] in, This is the satellite feature matrix corresponding to N satellites at the current time. The temporal convolution kernel parameter matrix, This is the time-feature transformation matrix.

[0050] The above technical solution integrates a high-level representation of satellite state with time-specific features into the satellite feature prediction matrix. It uses the time feature matrix to learn the temporal patterns of satellite motion and obtain topological prediction information, providing a data foundation for subsequent parameter adjustments.

[0051] Please see Figure 4 In some embodiments, the low-Earth orbit (LEO) satellite information includes Doppler frequency shift. Multi-source data fusion processing is performed based on the satellite feature prediction matrix, LEO satellite information, positioning system observation information, and a preset Kalman filter to obtain the floating-point result corresponding to the current location, including: S1031. Based on the satellite feature prediction matrix and Doppler frequency shift, update the noise matrix and the constraint matrix corresponding to the observation equation in the preset Kalman filter to obtain the updated observation equation and the updated constraint matrix. S1032. Update the preset Kalman filter based on the updated observation equation and the updated constraint matrix to obtain the updated preset Kalman filter; S1033. Multi-source data fusion processing is performed based on low-orbit satellite information, positioning system observation information, and the updated preset Kalman filter to obtain floating-point results.

[0052] The Doppler constraint terms introduced in the preset Kalman filter are dynamically adjusted based on the satellite feature prediction matrix, and adaptive adjustment is performed to obtain the updated preset Kalman filter.

[0053] In some embodiments, the formula for the actual measured Doppler frequency shift z is as follows:

[0054] Let v be the Doppler frequency shift observation under ideal conditions with no noise, v be the observation noise, v have a mean of 0, and v have a covariance matrix of R. Let be the original frequency of the carrier signal, i.e., the transmission frequency, and c be the speed of light. Let be the satellite's velocity vector, representing the satellite's instantaneous velocity at a given moment. Let be the velocity vector of the receiver to be located. It is a three-dimensional vector that represents the instantaneous velocity of the receiver at a certain moment. The component in the line-of-sight direction is the signal frequency number caused by the relative motion between the satellite and the receiver to be positioned. The greater the component of relative velocity in the line-of-sight direction, the greater the frequency shift. LEO satellites have significant Doppler frequency shifts due to their low orbital altitude and high relative velocity.

[0055] The pre-defined Kalman filter model with Doppler constraint terms is as follows:

[0056] in, This is the solution result at the current moment. This is the solution result from the previous cycle. G is the state transition matrix based on physical laws, and G is the noise driving matrix. Let B be a random process noise vector, and let B be a control constraint term. These are Doppler frequency shift observations.

[0057] In the above technical solution, the satellite prediction feature matrix is ​​not directly used to modify the state transition matrix. Instead of using core dynamic model parameters such as the noise-driven matrix G, the filter parameters are indirectly "adaptively adjusted" by influencing the confidence level and usage of constraint terms that affect the Doppler frequency shift. The Doppler frequency shift observations from low-Earth orbit satellites, weighted by "confidence level", are introduced into the state equation as a strong physical constraint, effectively correcting the errors generated during initial state prediction, especially the errors caused by inaccurate dynamic models, and significantly improving the accuracy and robustness of state estimation, i.e., positioning results.

[0058] Please see Figure 5 In some embodiments, positioning calculations are performed based on floating-point results, low-orbit satellite information, and a preset ambiguity fixed model to obtain target positioning results, including: S1041. Determine the solution state characteristic parameters based on low-Earth orbit satellite information. The solution state characteristic parameters are used to characterize the signal quality characteristics of multiple low-Earth orbit satellites. S1042. Determine the test threshold based on floating-point results and solution state characteristic parameters; S1043. Based on the floating-point results, perform evaluation processing to obtain a candidate set of integer solutions; S1044. Determine the target localization result based on the test threshold and the integer solution candidate set.

[0059] The calculation of state characteristic parameters includes, but is not limited to, the Position Dilution of Precision (PDOP) and Signal-to-Noise Ratio (SNR) based on time-series arrangement. The trace of the position covariance matrix is ​​calculated using the direction matrix formed by the satellite and the receiver to be positioned. PDOP is the amplification factor of the three-dimensional position error. SNR is the ratio of signal power to noise power; in satellite navigation, it usually refers to the ratio of the received carrier signal strength to the noise strength. The floating-point results are evaluated using a standard ambiguity fixing algorithm to obtain an integer candidate set. For example, the ambiguity fixing algorithm can be the Least-squares AM Biguity Decorrelation Adjustment (LAMBDA) algorithm. A verification threshold obtained through machine learning is used for validation to obtain successfully fixed candidate values. These successfully fixed candidate values ​​are then substituted back into the updated observation equation for position solving to obtain the target positioning result.

[0060] The above technical solution improves the speed and efficiency of ambiguity fixing by adaptively adjusting the inspection threshold through machine learning.

[0061] Please see Figure 6 In some embodiments, the floating-point result includes ambiguity parameters and the corresponding residual sequence. The test threshold is determined based on the floating-point result and the solution state characteristic parameters, including: S10421. Local features are extracted based on the residual sequence corresponding to the fuzziness parameter to obtain local features; S10422. Based on ambiguity parameters and solved state characteristic parameters, perform time series characteristic analysis to obtain time series characteristics; S10423. Based on local features and temporal features, perform fusion processing to obtain the target fusion features; S10424. Perform threshold adaptive mapping processing on the target fusion features to obtain the test threshold.

[0062] Convolution operations are performed on the floating-point results and the CNN channels in the preset ambiguity fixed model to obtain local features Feature_CNN. Based on the temporally arranged PDOP values ​​and the floating-point unambiguity parameter N_ambiguity, the temporal features h_t are obtained through the LSTM channels in the preset ambiguity fixed model. The local features Feature_CNN extracted by the CNN channels and the temporal features h_t extracted by the LSTM channels are concatenated and then input into a fully connected layer for fusion processing to obtain an adaptive test threshold.

[0063] In some embodiments, the CNN channel is input with a recent ambiguity residual sequence N_residual, and a one-dimensional convolution operation is performed using a kernel: Feature_CNN = ReLU(K * N_residual + b), where K is the experimental coefficient and b is the experimental constant, to obtain the local feature Feature_CNN. The LSTM channel is input with a temporally arranged sequence of states (PDOP, N_ambiguity), and processed by an LSTM network h_t = LSTM(h_{t-1}, [PDOP, N_ambiguity]) to generate a temporal feature h_t. The local feature Feature_CNN extracted by the CNN channel and the temporal feature h_t extracted by the LSTM channel are concatenated [Feature_CNN ⊕ h_t]. The concatenated fused feature is then input into a fully connected layer W_o and a sigmoid function, outputting an adaptive test threshold τ, calculated as follows: τ = 1.5 + 1.5 × sigmoid( W_o [Feature_CNN ⊕ h_t] ) The above technical solutions extract local features from the residual sequence and capture temporal dependencies from ambiguity parameters and solution state feature parameters, thereby achieving adaptive dynamic adjustment of the inspection threshold and improving the ambiguity fixing speed.

[0064] Please see Figure 7 In some embodiments, the integer candidate set includes optimal candidate solutions and suboptimal candidate solutions. The target localization result is determined based on a test threshold and the integer candidate solution set, including: S10441. Determine the target ratio based on the optimal candidate solution and the second-best candidate solution; S10442. Determine fixed results based on test thresholds and target ratios; S10443. If the fixation result is successful, the position is solved based on the optimal candidate solution and the updated observation equation to obtain the target positioning result.

[0065] Calculate the target ratio (Ratio) between the best and second-best candidate solutions in the integer candidate set. Compare the calculated Ratio with a dynamic verification threshold τ generated by machine learning. If Ratio > τ, the fixation is considered successful; otherwise, the fixation fails. If the fixation is successful, substitute the best candidate solution back into the updated observation equation for position solving to obtain the target localization result.

[0066] In some embodiments, after substituting the ambiguity N of the fixed optimal candidate solution into the observation equation, the unknown to be solved in the equation is the position coordinate of the receiver to be located. By solving the equation using methods such as the least squares method, the target positioning result at the centimeter or even millimeter level, which eliminates integer ambiguity error, can be obtained.

[0067] The above technical solution uses a dynamic test threshold to fix ambiguity, obtains a precise fixed solution, and outputs accurate target positioning results, thereby improving the efficiency and accuracy of positioning.

[0068] Please see Figure 8 According to a second aspect of this disclosure, a positioning device is provided for use in a navigation system, the navigation system including multiple low-Earth orbit (LEO) satellites, the LEO satellites being deployed in a hybrid polar-orbiting and non-polar-orbiting configuration, the device comprising: Information acquisition module 10 is used to acquire multi-source observation data corresponding to the current time. The multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites. The topology modeling module 20 is used to perform topology modeling processing based on low-orbit satellite information and a preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step of the current time step. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on low-orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on low-orbit satellite information. The floating-point number calculation module 30 is used to perform multi-source data fusion processing based on satellite feature prediction matrix, low-orbit satellite information, positioning system observation information and preset Kalman filter to obtain the floating-point number result corresponding to the current position. The positioning result determination module 40 is used to perform positioning calculation processing based on floating-point results, low-orbit satellite information and preset ambiguity fixed model to obtain the target positioning result.

[0069] According to a third aspect of this disclosure, a navigation system is provided, comprising a plurality of low-Earth orbit satellites and a positioning device as described in the second aspect, wherein the plurality of low-Earth orbit satellites are deployed in a hybrid configuration of polar and non-polar orbits.

[0070] In some embodiments, the functions or modules of the apparatus provided in this disclosure can be used to perform the methods described in the above method embodiments. The specific implementation can be referred to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.

[0071] This application provides a positioning device, which can be a terminal or a server. The positioning device includes a processor and a memory. The memory stores at least one instruction or at least one program. The at least one instruction or at least one program is loaded and executed by the processor to implement the positioning method provided in the above method embodiments.

[0072] Memory is used to store software programs and modules. The processor executes these stored software programs and modules to perform various functional applications and data processing. Memory can primarily consist of a program storage area and a data storage area. The program storage area stores the operating system, application programs required for functionality, etc.; the data storage area stores data created based on device usage, etc. Furthermore, memory can include high-speed random access memory (RAM) and non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, memory can also include a memory controller to provide the processor with access to the memory.

[0073] The methods and embodiments provided in this application can be executed in electronic devices such as mobile terminals, computer terminals, servers, or similar computing devices. Figure 9 This is a hardware structure block diagram of an electronic device using a positioning method provided in an embodiment of this application. For example... Figure 9As shown, the electronic device 900 can vary considerably due to differences in configuration or performance. It may include one or more Central Processing Units (CPUs) 910 (CPUs 910 may include, but are not limited to, microprocessors such as MCUs or programmable logic devices such as FPGAs), a memory 930 for storing data, and one or more storage media 920 (e.g., one or more mass storage devices) for storing application programs 923 or data 922. The memory 930 and storage media 920 may be temporary or persistent storage. The program stored in the storage media 920 may include one or more modules, each module may include a series of instruction operations on the electronic device. Furthermore, the CPU 910 may be configured to communicate with the storage media 920 and execute the series of instruction operations in the storage media 920 on the electronic device 900. Electronic device 900 may also include one or more power supplies 960, one or more wired or wireless network interfaces 950, one or more input / output interfaces 940, and / or one or more operating systems 921, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc.

[0074] The input / output interface 940 can be used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the electronic device 900. In one example, the input / output interface 940 includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the input / output interface 940 may be a radio frequency (RF) module used for wireless communication with the Internet.

[0075] Those skilled in the art will understand that Figure 9 The structure shown is for illustrative purposes only and does not limit the structure of the electronic device described above. For example, the electronic device 900 may also include... Figure 9 The more or fewer components shown, or having the same Figure 9 The different configurations shown.

[0076] Embodiments of this application also provide a computer-readable storage medium, which can be disposed in an electronic device to store at least one instruction or at least one program related to implementing a positioning method in the method embodiments. The at least one instruction or the at least one program is loaded and executed by the processor to implement the positioning method provided in the above method embodiments.

[0077] Optionally, in this embodiment, the storage medium may be located at at least one of the multiple network servers in a computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0078] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations described above.

[0079] As can be seen from the embodiments of the positioning method, apparatus, device, terminal, server, storage medium, or computer program provided in this application, the navigation system of this application includes multiple low-Earth orbit (LEO) satellites. The LEO satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration. Based on the original polar-orbiting configuration, a polar-orbiting plus inclined orbit deployment method is adopted to improve the coverage density in mid-latitude and longitude regions. Multi-source observation data corresponding to the current moment is acquired. This multi-source observation data includes LEO satellite information corresponding to multiple LEO satellites and positioning system observation information. By adding inclined orbit satellites, the spatial geometric distribution diversity of satellites is improved from a physical perspective, providing richer and more stable LEO satellite information for downstream calculations. Based on low-Earth orbit (LEO) satellite information and a pre-defined dynamic prediction model, topology modeling is performed to obtain the satellite feature prediction matrix corresponding to the next time step. This matrix is ​​a fusion of spatial and temporal feature matrices. The spatial feature matrix is ​​obtained by extracting spatial features from LEO satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features from LEO satellite information. The spatial feature matrix captures spatial dependencies, while the temporal feature matrix learns the temporal patterns of satellite motion, resulting in a high-level representation of satellite state that integrates spatiotemporal features. Multi-source data fusion processing is then performed based on the satellite feature prediction matrix, LEO satellite information, positioning system observation information, and a pre-defined Kalman filter to obtain a floating-point result corresponding to the current position. This multi-source data fusion processing addresses the issue of unstable observation data quality, reducing positioning errors. Finally, positioning calculation is performed based on the floating-point result, LEO satellite information, and a pre-defined ambiguity fixing model to obtain the target positioning result. The pre-defined ambiguity fixing model solves the problem of low ambiguity fixing efficiency, improving positioning efficiency and accuracy.

[0080] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired results. Additionally, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are also possible or may be advantageous.

[0081] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the device, equipment, and storage medium embodiments are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0082] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware, or by a program instructing the relevant hardware to implement them. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0083] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A positioning method, characterized in that, The method is applied to a navigation system comprising multiple low-Earth orbit (LEO) satellites, wherein the LEO satellites are deployed in a hybrid polar-orbiting and non-polar-orbiting configuration. Acquire multi-source observation data corresponding to the current moment, wherein the multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites; Based on the low-Earth orbit satellite information and the preset dynamic prediction model, topology modeling is performed to obtain the satellite feature prediction matrix corresponding to the next time step of the current time step. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on the low-Earth orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on the low-Earth orbit satellite information. Based on the satellite feature prediction matrix, the low-orbit satellite information, the positioning system observation information, and the preset Kalman filter, multi-source data fusion processing is performed to obtain the floating-point result corresponding to the current position; Based on the floating-point result, the low-orbit satellite information, and the preset ambiguity fixed model, the positioning calculation is performed to obtain the target positioning result.

2. The method according to claim 1, characterized in that, The topology modeling process based on the low-orbit satellite information and the preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step at the current time step includes: Based on the low-Earth orbit satellite information, a satellite feature matrix corresponding to the current time is constructed. The satellite feature matrix is ​​used to characterize the state features corresponding to the multiple low-Earth orbit satellites. A satellite adjacency matrix is ​​constructed based on the low-Earth orbit satellite information. The satellite adjacency matrix is ​​used to characterize the spatial dependencies between the multiple low-Earth orbit satellites. A weighted aggregation process is performed based on the satellite adjacency matrix and the satellite feature matrix to obtain an initial spatial fusion matrix; Based on the initial spatial fusion matrix and the spatial feature transformation matrix, feature transformation processing is performed to obtain the spatial feature matrix.

3. The method according to claim 2, characterized in that, The method further includes: The initial time fusion matrix is ​​obtained by weighted aggregation of the temporal convolution kernel parameter matrix of the preset dynamic prediction model and the satellite feature matrix. Based on the initial time fusion matrix and the time feature transformation matrix, feature transformation processing is performed to obtain the time feature matrix.

4. The method according to claim 1, characterized in that, The low-Earth orbit (LEO) satellite information includes Doppler frequency shift. The multi-source data fusion processing based on the satellite feature prediction matrix, the LEO satellite information, the positioning system observation information, and a preset Kalman filter to obtain the floating-point result corresponding to the current position includes: Based on the satellite feature prediction matrix and the Doppler frequency shift, update the noise matrix corresponding to the observation equation and the constraint matrix corresponding to the observation equation in the preset Kalman filter to obtain the updated observation equation and the updated constraint matrix; The preset Kalman filter is updated based on the updated observation equation and the updated constraint matrix to obtain the updated preset Kalman filter; The floating-point result is obtained by performing multi-source data fusion processing based on the low-orbit satellite information, the positioning system observation information, and the updated preset Kalman filter.

5. The method according to claim 1, characterized in that, The positioning calculation based on the floating-point result, the low-orbit satellite information, and the preset ambiguity fixed model yields the target positioning result, including: Based on the low-Earth orbit satellite information, the solution state characteristic parameters are determined, and the solution state characteristic parameters are used to characterize the signal quality characteristics corresponding to multiple low-Earth orbit satellites. The test threshold is determined based on the floating-point result and the solution state characteristic parameters; Based on the floating-point results, an evaluation process is performed to obtain a candidate set of integer solutions; The target localization result is determined based on the test threshold and the integer solution candidate set.

6. The method according to claim 5, characterized in that, The floating-point result includes ambiguity parameters and the residual sequence corresponding to the ambiguity parameters. Determining the test threshold based on the floating-point result and the solution state feature parameters includes: Local features are extracted based on the residual sequence corresponding to the ambiguity parameters to obtain local features. Based on the ambiguity parameters and the solution state characteristic parameters, time series characteristic analysis is performed to obtain time series characteristics; The target fusion feature is obtained by fusing the local features and the temporal features. The target fusion features are subjected to threshold adaptive mapping to obtain the test threshold.

7. The method according to claim 5, characterized in that, The integer candidate set includes optimal candidate solutions and suboptimal candidate solutions. Determining the target localization result based on the test threshold and the integer candidate solution set includes: The target ratio is determined based on the optimal candidate solution and the second-best candidate solution; A fixed result is determined based on the test threshold and the target ratio; If the fixation result is successful, the position is solved based on the optimal candidate solution and the updated observation equation to obtain the target positioning result.

8. A positioning device, characterized in that, The device is applied to a navigation system, which includes multiple low-Earth orbit (LEO) satellites deployed in a hybrid polar-orbiting and non-polar-orbiting configuration. The information acquisition module is used to acquire multi-source observation data corresponding to the current moment. The multi-source observation data includes low-orbit satellite information and positioning system observation information corresponding to multiple low-orbit satellites. The topology modeling module is used to perform topology modeling processing based on the low-Earth orbit satellite information and a preset dynamic prediction model to obtain the satellite feature prediction matrix corresponding to the next time step of the current time step. The satellite feature prediction matrix is ​​a matrix obtained by fusing spatial feature matrix and temporal feature matrix. The spatial feature matrix is ​​obtained by extracting spatial features based on the low-Earth orbit satellite information, and the temporal feature matrix is ​​obtained by extracting temporal features based on the low-Earth orbit satellite information. The floating-point number calculation module is used to perform multi-source data fusion processing based on the satellite feature prediction matrix, the low-orbit satellite information, the positioning system observation information, and the preset Kalman filter to obtain the floating-point number result corresponding to the current position; The positioning result determination module is used to perform positioning calculation processing based on the floating-point number result, the low-orbit satellite information and the preset ambiguity fixed model to obtain the target positioning result.

9. A navigation system, characterized in that, The navigation system includes multiple low-Earth orbit satellites and the positioning device as described in claim 8, wherein the multiple low-Earth orbit satellites are deployed in a hybrid configuration of polar orbit and non-polar orbit.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction or at least one program, which is loaded and executed by a processor to implement the positioning method as described in any one of claims 1-7.

Images