Multi-path effect detection method based on artificial intelligence

Abstract

A multi-path effect detection method based on artificial intelligence, comprising: determining the degree of correlation between a multi-path error and different feature values on the basis of historical observation data (101); acquiring real-time observation data and GIS data sent by a satellite (102); and inputting the real-time observation data, the GIS data, and the degree of correlation between the multi-path error and different feature values into a first model, and determining the degree of impact of the multi-path effect on the real-time observation data (103). According to the method, in one aspect, the degree of correlation between the multi-path error and different feature values is determined on the basis of the historical observation data, so that the features affecting the multi-path error can be accurately determined, and further, the multi-path error can be accurately and effectively detected; in another aspect, on the basis of the GIS data, environmental features possibly causing the multi-path effect can be accurately identified; and in another further aspect, the first model is used to detect the multi-path effect, so that complex temporal and spatial characteristics in multi-path can be covered, achieving accurate and effective detection of the multi-path error.

Description

A multipath effect detection method based on artificial intelligence Technical Field

[0001] The present application relates to the technical field of global navigation satellite systems, and in particular to a multipath effect detection method based on artificial intelligence. Background Art

[0002] With the rapid development of multi-constellation GNSS systems, such as the Global Positioning System (GPS), Beidou Navigation Satellite System (BDS), GLONASS, and Galileo, navigation and positioning accuracy has significantly improved. To obtain reliable and accurate positioning results, it is essential to address the associated errors. Multipath error is difficult to eliminate using traditional error processing methods, and there is a lack of suitable theoretical models to describe it. Multipath error has become one of the main sources of error affecting high-precision GNSS data processing. Multipath refers to the phenomenon in GNSS measurements where signals reflected or diffracted by obstructions near the station enter the receiver antenna, causing the observed value to deviate from the true value.

[0003] Since multipath errors cannot usually be completely eliminated directly, they often need to be detected first, and then significant multipath errors need to be suppressed. Currently, the detection of GNSS multipath errors mainly adopts the following methods. First, pseudorange multipath combination can be used. This method is essentially a geometry-free and ionosphere-free function model. However, pseudorange multipath combination can only reflect the fluctuations of pseudorange multipath and cannot accurately understand its multipath errors. Secondly, the double difference residual method can be used to evaluate phase multipath, but the main drawback of this method is that it is only effective on short baselines. Finally, the relationship between data eigenvalues ​​and multipath can be utilized. Compared with the first two methods, this method can accurately and reliably detect phase multipath errors. Due to the complex temporal and spatial characteristics of multipath, this brings certain challenges and difficulties to the real-time multipath modeling and high-precision prediction of GNSS systems.

[0004] Summary of the Invention

[0005] The present application provides an artificial intelligence-based multipath effect detection method that can accurately and effectively detect multipath errors.

[0006] In a first aspect, an embodiment of the present application provides a multipath effect detection method based on artificial intelligence, which can be executed by a multipath effect detection device based on artificial intelligence, and the multipath effect detection device based on artificial intelligence can be a terminal device or a module for a terminal device, or a server or a module for a server. The present application does not limit the execution subject of the method. The method includes: determining the degree of correlation between multipath error and different eigenvalues ​​based on historical observation data; obtaining real-time observation data and geographic information system (GIS) data sent by satellites; inputting the real-time observation data, the GIS data and the degree of correlation between the multipath error and different eigenvalues ​​into the first model to determine the degree of influence of the multipath effect on the real-time observation data; wherein the first model is trained based on the historical observation data, the GIS data and the degree of correlation between the multipath error and different eigenvalues.

[0007] The above scheme, on the one hand, determines the degree of correlation between multipath error and different eigenvalues ​​based on historical observation data, can accurately determine the characteristics that affect the multipath error, and further, can accurately and effectively detect the multipath error; on the other hand, based on GIS data, it can accurately identify environmental features that may cause multipath effects, such as buildings, vegetation, terrain and other information; on the other hand, using the first model to detect multipath effects can cover the complex temporal and spatial characteristics of multipath, and can accurately and effectively detect multipath errors.

[0008] In a possible implementation method, an observation value residual and multiple eigenvalues ​​are determined based on the historical observation data; wherein the observation value residual is used to indicate a multipath error; the observation value residual includes a single-difference pseudorange and a phase residual between different stations; the eigenvalue includes one or more of the following: satellite elevation angle, satellite azimuth angle, inter-frequency differential carrier-to-noise ratio, number of observed satellites, phase consistency, pseudorange consistency, pseudorange rate consistency, phase-frequency differential value, or time single-difference ambiguity; and the degree of correlation between each eigenvalue and the observation value residual is determined.

[0009] The above scheme determines the degree of correlation between the multipath error and different characteristic values ​​based on historical observation data, can accurately determine the characteristics that affect the multipath error, and further, can accurately and effectively detect the multipath error.

[0010] In a possible implementation method, the real-time observation data is stored in a first database, and the first database is updated.

[0011] The above scheme updates the first database in real time, making the data in the first database real-time and reliable, and as the data in the first database increases, the accuracy of the first model, the second model and the correlation model will also increase accordingly, further, it can achieve accurate and effective detection of multipath errors.

[0012] In one possible implementation method, the first model is periodically trained based on the updated first database.

[0013] The above solution updates the first database in real time, making the data in the first database real-time and reliable. As the data in the first database increases, the accuracy of the first model will also increase. Furthermore, accurate and effective detection of multipath errors can be achieved.

[0014] In one possible implementation method, a correlation model is periodically trained based on an updated first database; the correlation model is used to determine the degree of correlation between multipath errors and different eigenvalues; and determining the degree of correlation between multipath errors and different eigenvalues ​​based on historical observation data includes: inputting the historical observation data into the correlation model to determine the degree of correlation between multipath errors and different eigenvalues.

[0015] The above solution updates the first database in real time, making the data in the first database real-time and reliable. As the data in the first database increases, the accuracy of the correlation model will also increase, making it possible to accurately determine the correlation between multipath errors and different eigenvalues.

[0016] In one possible implementation method, observation data within a first time period sent by the satellite and GIS data within the first time period are obtained; the observation data within the first time period and the GIS data within the first time period are input into a second model to predict the degree of influence of the multipath effect after the first time period.

[0017] The above solution can predict the multipath effect in the future time period based on the observation data in the first time period, and can achieve timely and accurate processing of the multipath effect, making positioning more accurate.

[0018] In one possible implementation method, historical observation data is divided into multiple data to be processed in the first time period according to the second time period; multiple signal characteristic values ​​are determined from the multiple data to be processed; wherein each signal characteristic value is used to indicate the degree of influence of multipath error on the corresponding data to be processed; the correlation relationship between each signal characteristic value and environmental information is determined, and the environmental information includes time information and spatial information; the time information is used to indicate the periodic change law of the degree of influence of multipath error on the data to be processed; the spatial information is used to indicate the influence of GIS data on the data to be processed; and the second model is trained based on the correlation relationship between each signal characteristic value and the environmental information.

[0019] In the above scheme, the second model trained by the above method has time and space information, so it can accurately determine the time factors and space factors that affect the multipath effect, and then can achieve timely and accurate processing of the multipath effect, making positioning more accurate.

[0020] In a possible implementation method, based on the historical observation data, double-difference residuals of the historical observation data are obtained through a double-difference positioning mode of inter-station single difference and inter-satellite single difference; wherein the inter-station single difference refers to the differential processing of observation data between a mobile station and a reference station; the inter-satellite single difference refers to the differential processing of observation data between different satellites; the double-difference residual refers to the residual after inter-station single difference and inter-satellite single difference; based on the double-difference residual, the observation value residual is determined; wherein the observation value residual refers to the single-difference pseudorange and phase residual between the same satellite and different stations.

[0021] In this scheme, other types of unmodeled errors, such as residual ionospheric and tropospheric delays, can be considered completely eliminated. Especially when the baseline length is sufficiently short, pseudorange and phase residuals can largely be considered multipath corrections. Because this method is based on a single satellite, it is necessary to convert the double-difference multipath corrections (double-difference residuals) into single-difference corrections between receivers. This enables accurate determination of observation residuals and, further, the characteristics that influence multipath errors.

[0022] In one possible implementation method, the method for determining phase consistency includes the following processes: obtaining the phases of the first epoch and the second epoch respectively from the historical observation data; wherein the first epoch is the previous epoch of the second epoch; predicting the phase of the second epoch based on the phase of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and carrier wavelengths of different frequencies; determining the phase consistency of the second epoch based on the phase of the second epoch and the predicted phase of the second epoch; wherein the phase consistency is inversely correlated with the degree of influence of multipath error; the method for determining pseudorange consistency includes the following processes: obtaining the pseudoranges of the first epoch and the second epoch respectively from the historical observation data; predicting the pseudorange of the second epoch based on the pseudorange of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and carrier wavelengths of different frequencies; determining the pseudorange consistency of the second epoch based on the pseudorange of the second epoch and the predicted pseudorange of the second epoch; wherein the pseudorange consistency is inversely correlated with the degree of influence of multipath error.

[0023] The above scheme can accurately determine characteristics such as phase consistency and pseudorange consistency.

[0024] In one possible implementation method, a first pseudorange rate between the first epoch and the second epoch is determined from the historical observation data; a second pseudorange rate is obtained based on Doppler observations; and pseudorange rate consistency is determined based on the first pseudorange rate and the second pseudorange rate; wherein the pseudorange rate consistency is inversely correlated with the degree of influence of multipath error.

[0025] The above scheme can accurately determine the consistency characteristics of the pseudorange rate.

[0026] In a second aspect, an embodiment of the present application provides an artificial intelligence-based multipath effect detection device, comprising: an acquisition unit and a processing unit. The acquisition unit is configured to acquire real-time observation data and geographic information system (GIS) data transmitted by a satellite; the processing unit is configured to determine the degree of correlation between multipath error and different eigenvalues ​​based on historical observation data; the real-time observation data, the GIS data, and the degree of correlation between the multipath error and different eigenvalues ​​are input into a first model to determine the degree of influence of the multipath effect on the real-time observation data; wherein the first model is trained based on the historical observation data, the GIS data, and the degree of correlation between the multipath error and different eigenvalues.

[0027] In one possible implementation method, the processing unit is used to determine an observation value residual and multiple eigenvalues ​​based on the historical observation data; wherein the observation value residual is used to indicate a multipath error; the observation value residual includes a single-difference pseudorange and a phase residual between different stations; the eigenvalue includes one or more of the following: satellite elevation angle, satellite azimuth angle, inter-frequency differential carrier-to-noise ratio, number of observed satellites, phase consistency, pseudorange consistency, pseudorange rate consistency, phase-frequency differential value, or time single-difference ambiguity; and determine the degree of correlation between each eigenvalue and the observation value residual.

[0028] In a possible implementation method, the processing unit is configured to store the real-time observation data in a first database and update the first database.

[0029] In a possible implementation method, the processing unit is used to periodically train the first model based on the updated first database.

[0030] In one possible implementation method, the processing unit is used to periodically train the correlation model based on the updated first database; the correlation model is used to determine the degree of correlation between the multipath error and different eigenvalues; the historical observation data is input into the correlation model to determine the degree of correlation between the multipath error and different eigenvalues.

[0031] In one possible implementation method, the acquisition unit is used to obtain observation data within a first time period sent by the satellite and GIS data within the first time period; the processing unit is used to input the observation data within the first time period and the GIS data within the first time period into a second model to predict the degree of influence of the multipath effect after the first time period.

[0032] In one possible implementation method, the processing unit is used to divide the historical observation data into multiple data to be processed in the first time period according to the second time period; determine multiple signal characteristic values ​​from the multiple data to be processed; wherein each signal characteristic value is used to indicate the degree of influence of multipath error on the corresponding data to be processed; determine the correlation between each signal characteristic value and environmental information, wherein the environmental information includes time information and spatial information; the time information is used to indicate the periodic change law of the degree of influence of multipath error on the data to be processed; the spatial information is used to indicate the influence of GIS data on the data to be processed; and train the second model based on the correlation between each signal characteristic value and the environmental information.

[0033] In a possible implementation method, the processing unit is used to obtain double-difference residuals of the historical observation data through a double-difference positioning mode of inter-station single difference and inter-satellite single difference based on the historical observation data; wherein the inter-station single difference refers to differential processing of observation data between a mobile station and a reference station; the inter-satellite single difference refers to differential processing of observation data between different satellites; the double-difference residual refers to the residual after inter-station single difference and inter-satellite single difference; based on the double-difference residual, the observation value residual is determined; wherein the observation value residual refers to the single-difference pseudorange and phase residual between the same satellite and different stations.

[0034] In one possible implementation method, the processing unit is used to obtain the phase of the first epoch and the second epoch from the historical observation data respectively; wherein the first epoch is the previous epoch of the second epoch; based on the phase of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies, predict the phase of the second epoch; based on the phase of the second epoch and the predicted phase of the second epoch, determine the phase consistency of the second epoch; wherein the phase consistency is inversely correlated with the degree of influence of multipath error; respectively obtain the pseudorange of the first epoch and the second epoch from the historical observation data; based on the pseudorange of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies, predict the pseudorange of the second epoch; based on the pseudorange of the second epoch and the predicted pseudorange of the second epoch, determine the pseudorange consistency of the second epoch; wherein the pseudorange consistency is inversely correlated with the degree of influence of multipath error.

[0035] In one possible implementation method, the processing unit is configured to determine a first pseudorange rate between the first epoch and the second epoch from the historical observation data; obtain a second pseudorange rate based on a Doppler observation value; and determine pseudorange rate consistency based on the first pseudorange rate and the second pseudorange rate; wherein the pseudorange rate consistency is inversely correlated with the degree of influence of multipath error.

[0036] In a third aspect, an embodiment of the present application further provides a computing device, including:

[0037] a memory for storing program instructions;

[0038] The processor is configured to call the program instructions stored in the memory and execute any method for implementing the first aspect according to the obtained program instructions.

[0039] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, in which computer-readable instructions are stored. When a computer reads and executes the computer-readable instructions, any method of the above-mentioned first aspect is implemented.

[0040] In a fifth aspect, an embodiment of the present application provides a computer program product, comprising a computer program executable by a computer device, wherein when the program is run on the computer device, the computer device executes any method for implementing the above-mentioned first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] FIG1 is a schematic diagram of a flow chart of a multipath effect detection method based on artificial intelligence provided in an embodiment of the present application;

[0042] FIG2 is a flow chart of a multi-path effect prediction method based on artificial intelligence provided in an embodiment of the present application;

[0043] FIG3 is a flow chart of a training process of a second model provided in an embodiment of the present application;

[0044] FIG4 is a schematic diagram of the structure of an artificial intelligence-based multipath effect detection system provided in an embodiment of the present application;

[0045] FIG5 is a schematic structural diagram of an artificial intelligence-based multipath effect detection device provided in an embodiment of the present application;

[0046] FIG6 is a schematic structural diagram of an artificial intelligence-based multipath effect detection device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] The following is an explanation of some of the professional terms used in this application.

[0048] Global Navigation Satellite System (GNSS): The Global Navigation Satellite System is an airborne radio navigation and positioning system that can provide users with all-weather 3D coordinates, velocity, and time information at any location on the Earth's surface or in near-Earth space.

[0049] Multipath effect: If a satellite signal reaches a receiver via a straight-line path, the signal received by the receiver is delayed by the satellite's transmitted signal and has the highest signal strength. During signal propagation, reflections from objects change the signal's direction, amplitude, polarization, and phase. These altered signals arrive at the receiver and overlap with the signal that reached the receiver via the straight-line path. This phenomenon is called multipath. Multipath is one of the main causes of interference with GPS measurement quality. Multipath is similar to the phenomenon of echo: while the receiver receives the direct signal from the satellite, it also receives satellite signals reflected by other objects. If changing the measurement location is not possible in GPS measurements, the main method to reduce multipath is to increase the satellite cutoff angle. However, this also blocks signals from satellites at low altitudes (i.e., those just rising above the horizon). Other options include adding chokes and path-stopping shields. However, multipath can only be reduced, not eliminated.

[0050] Pseudorange: A measurement used in the Global Navigation Satellite System (GNSS). It represents the distance obtained by multiplying the time it takes a satellite signal to be transmitted from the satellite and received by the receiver by the speed of light. For example, to determine its position, a Global Positioning System (GNSS) receiver determines the distances to several satellites and their positions at the time of transmission. Knowing the satellite's orbital parameters allows these positions to be calculated.

[0051] Ionospheric delay refers to the atmospheric layer between 50 and 1000 km above Earth's surface. Radiation from the sun and other celestial bodies causes intense ionization of gas molecules in the troposphere, resulting in a large number of free electrons and positive ions of equal density. The presence of charged particles affects the propagation of electromagnetic waves, causing changes in propagation speed and bending of the propagation path. This causes the product of the signal propagation time and the speed of light in a vacuum to differ from the geometric distance between the signal's origin and reception point. This deviation is known as ionospheric delay or ionospheric refraction error.

[0052] Tropospheric delay: In the GNSS field, tropospheric delay is a source of positioning error. Tropospheric delay in satellite navigation positioning generally refers to the signal delay incurred by electromagnetic wave signals as they pass through the non-ionized, neutral atmosphere below 50 km. In studying signal delay, we no longer subdivide this atmospheric layer into the troposphere and stratosphere (as in atmospheric science) and disregard the differences in their properties. Because 80% of delay occurs in the troposphere, we refer to signal delay occurring in this neutral atmosphere as tropospheric delay. The troposphere is the lower portion of the atmosphere and is non-dispersive for frequencies up to 15 GHz. In this medium, the phase and group velocities associated with GPS carrier waves and signal information (PRN codes and navigation data) on L1 and L2 are equally delayed relative to free-space propagation. This delay varies with the refractive index of the troposphere, which depends on local temperature, pressure, and relative humidity. If not compensated, the equivalent distance of this delay can range from about 2.4 m with the satellite at zenith and the user at sea level to about 25 m with the satellite at an elevation angle of about 5°.

[0053] Inter-frequency differential carrier-to-noise ratio: refers to the difference in carrier-to-noise ratio between different frequencies of the same satellite.

[0054] Ambiguity: It is the unknown number in a whole cycle corresponding to the first observed value of the phase difference between the carrier phase and the reference phase when measuring the carrier phase of the Global Positioning System (GNSS).

[0055] FIG1 is a schematic flow chart of an artificial intelligence-based multipath effect detection method provided in an embodiment of the present application. This method can be performed by an artificial intelligence-based multipath effect detection device, which can be a terminal device or a module for a terminal device, or a server or a module for a server. This application does not limit the execution entity of this method.

[0056] The method comprises the following steps:

[0057] Step 101: Determine the correlation between multipath error and different eigenvalues ​​based on historical observation data.

[0058] In one possible implementation, upon receiving observation data transmitted by a satellite, a receiver stores the data in a first database. This database stores a large amount of observation data from different time periods, referred to as historical observation data. This solution updates the first database in real time, ensuring that the data in the database is always up-to-date and reliable. As the data in the first database grows, the accuracy of the first model, the second model, and the correlation model also improves, enabling accurate and effective detection of multipath errors.

[0059] In one possible implementation, observation residuals and multiple eigenvalues ​​are determined based on the historical observation data; the observation residuals are used to indicate multipath error; the observation residuals include single-difference pseudoranges and phase residuals between different stations; the eigenvalues ​​include one or more of the following: satellite elevation angle, satellite azimuth angle, inter-frequency differential carrier-to-noise ratio, number of observed satellites, phase consistency, pseudorange consistency, pseudorange rate consistency, phase-frequency differential value, or time single-difference ambiguity; and the degree of correlation between each eigenvalue and the observation residual is determined. This solution, by determining the degree of correlation between multipath error and different eigenvalues ​​based on historical observation data, can accurately determine the characteristics that affect multipath error and, further, accurately and effectively detect multipath error.

[0060] In one possible implementation, double-difference residuals of the historical observation data are obtained using a double-difference positioning mode using inter-station single-difference and inter-satellite single-difference. The inter-station single-difference refers to the difference processing of observation data between a mobile station and a reference station; the inter-satellite single-difference refers to the difference processing of observation data between different satellites; the double-difference residuals refer to the residuals after inter-station single-difference and inter-satellite single-difference. Based on the double-difference residuals, observation residuals are determined. The observation residuals refer to single-difference pseudorange and phase residuals between the same satellite and different stations. In this scheme, other types of non-modeled errors, such as residual ionospheric and tropospheric delays, can be considered completely eliminated. Especially when the baseline length is sufficiently short, the pseudorange and phase residuals can largely be considered multipath corrections. Because this method is based on a single satellite, it is necessary to convert the double-difference multipath corrections (double-difference residuals) into single-difference corrections between receivers. It is possible to accurately determine the residual of the observation value, and further, it is possible to accurately determine the characteristics that affect the multipath error.

[0061] Step 102: Acquire real-time observation data and geographic information system (GIS) data sent by satellites.

[0062] In one possible implementation, geographic information system (GIS) data includes information on buildings, vegetation, terrain, etc. to help identify environmental features that may cause multipath effects.

[0063] Step 103: Input the real-time observation data, the GIS data, and the correlation between the multipath error and different eigenvalues ​​into the first model to determine the influence of the multipath effect on the real-time observation data.

[0064] The first model is trained based on the historical observation data, the GIS data, and the degree of correlation between the multipath error and different eigenvalues.

[0065] In a possible implementation method, based on the correlation between the multipath error and different eigenvalues, an eigenvalue with a higher correlation with the multipath error is selected and input into the first model.

[0066] In one possible implementation, the first model is trained based on historical observation data, GIS data, and the correlation between multipath errors and different eigenvalues. The algorithms used in the first model include random forests, attention mechanisms, neural networks, regression models, neural networks, support vector machines, and the like. This application does not place any restrictions on the algorithms used in the first model. The first model can be used to simulate signal propagation paths in real time.

[0067] In one possible implementation method, the real-time observation data is stored in a first database, and the first database is updated. A first model is periodically trained based on the updated first database. The parameters of the first model are updated in real time based on the updated first database to adapt to environmental changes and reduce the impact of multipath effects. The updated first model is then used to perform real-time verification to evaluate the accuracy and performance of the first model. This solution updates the first database in real time, making the data in the first database real-time and reliable. As the amount of data in the first database increases, the accuracy of the first model, the second model, and the correlation model also increases. Furthermore, accurate and effective detection of multipath errors can be achieved.

[0068] The above scheme, on the one hand, determines the degree of correlation between multipath error and different eigenvalues ​​based on historical observation data, can accurately determine the characteristics that affect the multipath error, and further, can accurately and effectively detect the multipath error; on the other hand, based on GIS data, it can accurately identify environmental features that may cause multipath effects, such as buildings, vegetation, terrain and other information; on the other hand, using the first model to detect multipath effects can cover the complex temporal and spatial characteristics of multipath, and can accurately and effectively detect multipath errors.

[0069] In one possible implementation method, in step 101 above, the correlation between the multipath error and different eigenvalues ​​is determined based on historical observation data, including: inputting the historical observation data into the correlation model to determine the correlation between the multipath error and different eigenvalues. The correlation model is used to determine the correlation between the multipath error and different eigenvalues; the correlation model is obtained by periodic training based on the updated first database; that is, after the first database is updated, the observation value residuals and multiple eigenvalues ​​determined by the historical observation data are input into the correlation model for training. Of course, the correlation model can also be trained periodically at certain time intervals, and this application does not limit this.

[0070] In one possible implementation method, the multipath effect and correlation analysis results, the correlation model, and the first model are stored to provide support for suppressing multipath in different application scenarios. The storage system can be the first database or other storage system, which is not limited in this application.

[0071] FIG2 is a flow chart of a multi-path effect prediction method based on artificial intelligence provided in an embodiment of the present application, the method comprising the following steps:

[0072] Step 201: Acquire observation data within a first time period and GIS data within the first time period sent by a satellite.

[0073] In one possible implementation method, the present application does not limit the first time period. For example, the first time period is 10 days, that is, the multipath effect of the next day or the next week can be predicted based on the observation data within the 10 days.

[0074] In a possible implementation method, the observation data in the first time period includes real-time observation data and / or historical observation data.

[0075] Step 202: input the observation data within the first time period and the GIS data within the first time period into the second model to predict the impact of the multipath effect after the first time period.

[0076] In one possible implementation method, the second model is trained based on multiple observation data within the first time period and GIS data within the first time period. The algorithms used by the second model include random forest, attention mechanism, neural network, regression model, neural network, support vector machine, etc. This application does not limit the algorithm of the second model.

[0077] The above solution can predict the multipath effect in the future time period based on the observation data in the first time period, and can achieve timely and accurate processing of the multipath effect, making positioning more accurate.

[0078] In one possible implementation method, the training process of the second model is shown in FIG3 , and the process includes the following steps:

[0079] Step 301: divide the historical observation data into a plurality of data to be processed in the first time period according to the second time period.

[0080] In one possible implementation method, the historical observation data is divided into multiple first time periods of data to be processed based on the second time period. For example, if the second time period is one month and the first time period is 10 days, the historical observation data for one month can be divided into multiple 10-day data to be processed, with each time period corresponding to each data to be processed separated by one day. This application does not limit the second time period, the first time period, or the division method.

[0081] Step 302: Determine a plurality of signal characteristic values ​​from the plurality of data to be processed.

[0082] Among them, each signal characteristic value is used to indicate the degree of influence of multipath error on the corresponding data to be processed.

[0083] In a possible implementation method, a plurality of data to be processed and different eigenvalues ​​corresponding to the plurality of data to be processed are input into a correlation model to determine the degree of correlation between the multipath error and the different eigenvalues.

[0084] Step 303: Determine the association relationship between each signal characteristic value and environmental information.

[0085] Among them, the environmental information includes time information and spatial information; the time information is used to indicate the periodic change law of the degree of influence of multipath error on the data to be processed; the spatial information is used to indicate the influence of GIS data on the data to be processed.

[0086] One possible implementation involves using spatiotemporal analysis techniques, such as time series analysis and spatial interpolation, to establish relationships between signal characteristics and environmental variables, taking into account how the signal characteristics vary across time and space. This means that if multipath significantly impacts the observed data at a given time point, then as the satellite rotates, the impact may also be significantly greater at the same time point after one rotation. This also applies to spatial influences. For example, if a lake is present at one observation site, the multipath effect significantly impacts the observed data. If a lake is also present at another observation site, the multipath effect may also significantly impact the observed data.

[0087] Step 304: train the second model based on the association between each signal feature value and the environmental information.

[0088] In one possible implementation method, the second model is optimized based on the multipath effect of the real-time observation data to improve the prediction accuracy.

[0089] In the above scheme, the second model trained by the above method has time and space information, so it can accurately determine the time factors and space factors that affect the multipath effect, and then can achieve timely and accurate processing of the multipath effect, making positioning more accurate.

[0090] In one possible implementation method, the construction process of the first model and the second model is as follows:

[0091] Data collection, including observation data and corresponding GIS data.

[0092] Data processing: pre-processing of observation data and GIS data. This application does not limit the pre-processing method.

[0093] Feature extraction: obtain the characteristic values ​​of observation data with strong correlation with multipath, extract key characteristic values ​​(such as building height, shape, vegetation density, etc.) from environmental data, and fully consider the influencing factors of multipath.

[0094] Build a model using machine learning or deep learning algorithms (such as regression models, neural networks, support vector machines, etc.), and select an appropriate model or improve the model based on the characteristics of the data.

[0095] Training models and real-time updates: Use historical data to train the model, use an independent validation dataset to verify the model's performance, and use real-time observation data to perform incremental learning on the model to ensure that it can accurately detect and predict multipath effects.

[0096] The prediction results are presented in a visual way and explanations of the prediction results are provided. This can help users understand the prediction of multipath effects and facilitate the handling of multipath effects.

[0097] In a possible implementation method, the observation value residual is determined according to the historical observation data, including: according to the historical observation data, obtaining the double difference residual of the historical observation data through a double difference positioning mode of inter-station single difference and inter-satellite single difference; wherein the inter-station single difference refers to the differential processing of the observation data between the mobile station and the reference station; the inter-satellite single difference refers to the differential processing of the observation data between different satellites; the double difference residual refers to the residual after the inter-station single difference and the inter-satellite single difference; according to the double difference residual, the observation value residual is determined; wherein the observation value residual refers to the single difference pseudorange and phase residual between the same satellite and different stations.

[0098] In a possible implementation method, the double-difference residuals of the historical observation data include pseudorange double differences and phase double differences, wherein the pseudorange double differences and phase double differences are determined by the following formulas (1) and (2):

[0099] Wherein, the superscripts s and g represent the reference satellite and the common-view satellite, respectively; the subscripts r and q represent the reference station and the mobile station, respectively; represents the double-difference pseudorange observation; represents the double-difference phase observation; Represents the double difference distance between the receiver and the satellite; stands for double-difference ionospheric delay; represents the double-difference tropospheric delay; represents the double-difference pseudorange multipath effect; represents the double-difference pseudorange observation noise; represents the double-difference integer ambiguity; represents the double-difference phase multipath effect; represents the double-difference phase observation noise; λ f Indicates the carrier wavelength of different frequencies f; Indicates that two differences are performed.

[0100] In one possible implementation method, after obtaining the double-difference residual, in order to more accurately describe the multipath error, the double-difference residual can be converted into a single-difference residual, where the conversion method is shown in the following formula (3):

[0101] Among them, A and B represent different observation stations, ω i Represents the weighting of the i-th satellite using the altitude angle weighting function, θ represents the corresponding altitude angle, and satisfies ω i =sin 2 (θ) and represents the double difference residual of satellites i and j at sites A and B; represents the single-difference residual of satellite i at sites A and B.

[0102] In one possible implementation method, the site where the receiver is set up for observation is called an observation station. The observation station includes a reference station and a mobile station, and the receiver is located at the mobile station.

[0103] In one possible implementation, the pseudorange and phase double-difference residuals can be considered completely eliminated, since other types of unmodeled errors, such as residual ionospheric and tropospheric delays, can be considered completely eliminated. Especially when the baseline length is sufficiently short, the pseudorange and phase residuals can be largely considered multipath corrections. Because this method is based on a single satellite, it is necessary to convert the double-difference multipath corrections (double-difference residuals) into single-difference corrections between receivers.

[0104] One possible implementation method involves obtaining the Dilution of Precision (DOP) and innovation vector during data processing. Typically, positioning error is closely related to the product of the DOP and the measurement error. Regions with strong multipath errors tend to have larger DOPs. The innovation vector is typically expressed as the difference between the actual measured value and the corresponding predicted value. The Kalman filter continuously updates the state estimate and incorporates innovation information between the measured and predicted values ​​to improve the accuracy of the estimated system state. It is worth noting that positioning accuracy is not only affected by ranging accuracy but also closely related to the spatial geometric distribution of the observed satellites. The DOP describes the combined effect of the relative geometric arrangement of satellites. As the number of satellites increases, the DOP decreases monotonically. Positioning accuracy in three-dimensional space is generally of greater interest. Therefore, the Position Dilution of Precision (PDOP) is often used to measure the spatial distribution of observed satellites. At present, some studies have used the correlation function of PDOP as the variance factor between systems, which simplifies the computational complexity of estimating the weights of different observation values ​​using the posterior variance. The most widely used models are the cosine function and exponential function based on PDOP.

[0105] In one possible implementation method, the method for determining phase consistency includes the following process: obtaining the phase of the first epoch and the second epoch from the historical observation data respectively; wherein the first epoch is the epoch before the second epoch; predicting the phase of the second epoch based on the phase of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and carrier wavelengths of different frequencies; determining the phase consistency of the second epoch based on the phase of the second epoch and the predicted phase of the second epoch; wherein the phase consistency is inversely correlated with the degree of influence of the multipath error.

[0106] In one possible implementation method, the method for predicting the phase of the second epoch is shown in the following formula (4):

[0107] Among them, L k,k-1 They represent the phase predicted at the next epoch k at epoch k-1; L k-1 is the phase at epoch k-1; D k and D k-1 They represent the Doppler observation values ​​at epochs k and k-1 respectively; λ represents the carrier wavelength of different frequencies; Δt represents the epoch interval.

[0108] In one possible implementation method, the method for determining phase consistency is shown in the following formula (5):

[0109] Where dL′ k They represent the phase consistency at epoch k; dL represents the time-difference phase observation value.

[0110] In one possible implementation method, the method for determining pseudorange consistency includes the following process: obtaining the pseudoranges of a first epoch and a second epoch from the historical observation data, respectively; predicting the pseudorange of the second epoch based on the pseudorange of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and carrier wavelengths of different frequencies; determining the pseudorange consistency of the second epoch based on the pseudorange of the second epoch and the predicted pseudorange of the second epoch; wherein the pseudorange consistency is inversely correlated with the degree of influence of multipath error.

[0111] In one possible implementation method, the method for predicting the pseudorange of the second epoch is shown in the following formula (6):

[0112] Among them, P k,k-1 They represent the pseudorange observation value predicted at the next epoch k at epoch k-1; P k-1 is the pseudorange observation value at epoch k-1; D k and D k-1 They represent the Doppler observation values ​​at epochs k and k-1 respectively; λ represents the carrier wavelength of different frequencies; Δt represents the epoch interval.

[0113] In one possible implementation method, the method for determining pseudorange consistency is shown in the following formula (7):

[0114] Where dP′ k They represent the pseudorange consistency at epoch k; dP represents the time-differenced pseudorange observation value.

[0115] In one possible implementation method, a method for determining pseudorange rate consistency includes the following process: determining a first pseudorange rate between the first epoch and the second epoch from the historical observation data; obtaining a second pseudorange rate based on Doppler observations; and determining pseudorange rate consistency based on the first pseudorange rate and the second pseudorange rate; wherein the pseudorange rate consistency is inversely correlated with the degree of influence of multipath error.

[0116] The above scheme can accurately determine characteristics such as phase consistency and pseudorange consistency.

[0117] In one possible implementation method, the method for determining the consistency of the pseudorange rate is shown in the following formulas (8) to (10):

[0118] Where Ω represents the pseudorange rate consistency; is the pseudorange rate; and is the pseudorange observation value of satellite s at the kth and k-1th epochs; is the pseudorange rate obtained by satellite s through Doppler observations.

[0119] The above scheme can accurately determine the consistency characteristics of the pseudorange rate.

[0120] In one possible implementation method, during the data processing process, the method for calculating the phase-frequency difference value of the observed data is shown in the following formula (11):

[0121] in, Indicates the double difference between the receiver and satellite phase hardware.

[0122] In one possible implementation method, during the data processing process, the method for calculating the time single-difference ambiguity of the observation data is shown in the following formula (12):

[0123] in, are the raw GNSS pseudorange and phase observations.

[0124] In one possible implementation method, the raw GNSS pseudorange and phase observations are calculated as shown in the following formulas (13) and (14):

[0125] The subscripts s, r, and f represent satellite, receiver, and frequency, respectively; are the raw GNSS pseudorange and phase observations, is the satellite-to-ground distance, δt r , δt s represent the satellite clock error and receiver clock error respectively, are the tropospheric delay and ionospheric delay, ξ r,f ,ξ s,f is the receiver and satellite pseudorange hardware bias, ζ r,f ,ζ s,f is the receiver and satellite phase hardware deviation, λ f for Upper phase ambiguity and carrier phase wavelength, denote the observation noise of pseudorange and phase respectively, represent the multipath effects of pseudorange and phase respectively.

[0126] In one possible implementation, because ξ r,f ,ξs,f ,ζ r,f ,ζ s,f These are all physical properties of the hardware and can be considered constants in a short period of time. The ionospheric delay is also a slowly changing error term. By performing inter-epoch differencing, we can remove those hardware delay terms and obtain the whole-cycle ambiguity, as shown in the following formula (15):

[0127] Where d is the time difference operator.

[0128] In one possible implementation method, and This reflects the change in ambiguity between two epochs. For LOS signals, cycle slips do not occur very frequently, but for NLOS signals, cycle slips occur frequently due to the rapid changes in reflection points and signal paths. For LOS signals, the time difference ambiguity is always within a certain range (1 meter), but for NLOS signals, the time difference ambiguity fluctuates greatly.

[0129] Based on the same technical concept, Figure 4 exemplarily shows a structural diagram of an artificial intelligence-based multipath effect detection system provided in an embodiment of the present application. As shown in Figure 4, the positioning model is used to process the observed GNSS data and determine the position of the receiver on the earth. There are many types of positioning models, and double-difference positioning is one of them.

[0130] In one possible implementation method, the correlation analysis results, the first model, and the second model in different application scenarios can be stored, and the observation environment and its multipath effect in different scenarios can be intelligently statistically analyzed and intelligently classified. Scenes with similar observation environments and similar multipath effects can be grouped into one category and stored. The stored data, parameters, and models can provide theoretical support for the multipath processing of other measurement points in this type of scene.

[0131] Based on the same technical concept, FIG5 exemplarily shows an artificial intelligence-based multipath effect detection device 500 provided in an embodiment of the present application. As shown in FIG5 , it includes: an acquisition unit 501 and a processing unit 502. The acquisition unit 501 is used to acquire real-time observation data and geographic information system GIS data sent by satellites; the processing unit 502 is used to determine the degree of correlation between multipath errors and different eigenvalues ​​based on historical observation data; the real-time observation data, the GIS data and the degree of correlation between the multipath errors and different eigenvalues ​​are input into the first model to determine the degree of influence of the multipath effect on the real-time observation data; wherein, the first model is trained based on the historical observation data, the GIS data and the degree of correlation between the multipath errors and different eigenvalues.

[0132] In one possible implementation method, the processing unit 502 is used to determine an observation value residual and multiple eigenvalues ​​based on the historical observation data; wherein the observation value residual is used to indicate a multipath error; the observation value residual includes a single-difference pseudorange and a phase residual between different stations; the eigenvalue includes one or more of the following: satellite altitude angle, satellite azimuth angle, inter-frequency differential carrier-to-noise ratio, number of observed satellites, phase consistency, pseudorange consistency, pseudorange rate consistency, phase-frequency differential value, or time single-difference ambiguity; and determine the degree of correlation between each eigenvalue and the observation value residual.

[0133] In a possible implementation method, the processing unit 502 is configured to store the real-time observation data in a first database and update the first database.

[0134] In a possible implementation method, the processing unit 502 is used to periodically train the first model based on the updated first database.

[0135] In one possible implementation method, the processing unit 502 is used to periodically train the correlation model based on the updated first database; the correlation model is used to determine the degree of correlation between the multipath error and different eigenvalues; the historical observation data is input into the correlation model to determine the degree of correlation between the multipath error and different eigenvalues.

[0136] In one possible implementation method, the acquisition unit 501 is used to obtain observation data within a first time period sent by the satellite and GIS data within the first time period; the processing unit 502 is used to input the observation data within the first time period and the GIS data within the first time period into a second model to predict the degree of influence of the multipath effect after the first time period.

[0137] In one possible implementation method, the processing unit 502 is used to divide the historical observation data into multiple data to be processed in the first time period according to the second time period; determine multiple signal characteristic values ​​from the multiple data to be processed; wherein each signal characteristic value is used to indicate the degree of influence of the multipath error on the corresponding data to be processed; determine the correlation between each signal characteristic value and environmental information, and the environmental information includes time information and spatial information; the time information is used to indicate the periodic change law of the degree of influence of the multipath error on the data to be processed; the spatial information is used to indicate the influence of GIS data on the data to be processed; and train the second model according to the correlation between each signal characteristic value and the environmental information.

[0138] In one possible implementation method, the processing unit 502 is used to obtain double-difference residuals of the historical observation data through a double-difference positioning mode of inter-station single difference and inter-satellite single difference based on the historical observation data; wherein the inter-station single difference refers to differential processing of observation data between a mobile station and a reference station; the inter-satellite single difference refers to differential processing of observation data between different satellites; the double-difference residual refers to the residual after inter-station single difference and inter-satellite single difference; based on the double-difference residual, the observation value residual is determined; wherein the observation value residual refers to the single-difference pseudorange and phase residual between the same satellite and different stations.

[0139] In one possible implementation method, the processing unit 502 is used to obtain the phase of the first epoch and the second epoch from the historical observation data, respectively; wherein the first epoch is the previous epoch of the second epoch; based on the phase of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies, predict the phase of the second epoch; based on the phase of the second epoch and the predicted phase of the second epoch, determine the phase consistency of the second epoch; wherein the phase consistency is inversely correlated with the degree of influence of multipath error; respectively obtain the pseudorange of the first epoch and the second epoch from the historical observation data; based on the pseudorange of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies, predict the pseudorange of the second epoch; based on the pseudorange of the second epoch and the predicted pseudorange of the second epoch, determine the pseudorange consistency of the second epoch; wherein the pseudorange consistency is inversely correlated with the degree of influence of multipath error.

[0140] In one possible implementation method, the processing unit 502 is configured to determine a first pseudorange rate between the first epoch and the second epoch from the historical observation data; obtain a second pseudorange rate based on a Doppler observation value; and determine pseudorange rate consistency based on the first pseudorange rate and the second pseudorange rate; wherein the pseudorange rate consistency is inversely correlated with the degree of influence of multipath error.

[0141] Based on the same technical concept, an embodiment of the present application provides an artificial intelligence-based multipath effect detection device 600. This multipath effect detection device 600 can be, for example, a computing device. As shown in Figure 6 , this multipath effect detection device 600 includes at least one processor 601 and a memory 602 connected to the at least one processor. The specific connection medium between processor 601 and memory 602 is not limited in this embodiment of the application. In Figure 6 , the connection between processor 601 and memory 602 is illustrated as a bus. Buses can be classified into address buses, data buses, control buses, and so on.

[0142] In an embodiment of the present application, the memory 602 stores instructions that can be executed by at least one processor 601. The at least one processor 601 can execute the above-mentioned artificial intelligence-based multipath effect detection method by executing the instructions stored in the memory 602.

[0143] Processor 601 serves as the control center of multipath effect detection device 600. It can connect various components of the computer device using various interfaces and circuits, and configure resources by running or executing instructions stored in memory 602 and accessing data stored in memory 602. Optionally, processor 601 may include one or more determination units. Processor 601 may integrate an application processor and a modem processor. The application processor primarily processes the operating system, user interface, and application programs, while the modem processor primarily handles wireless communications. It is understood that the modem processor may not be integrated into processor 601. In some embodiments, processor 601 and memory 602 may be implemented on the same chip. In some embodiments, they may also be implemented on separate chips.

[0144] The processor 601 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit (ASIC), a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in conjunction with the embodiments of the present application can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor.

[0145] The memory 602 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 602 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 602 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 602 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0146] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer-executable program. The computer-executable program is used to enable a computer to execute an artificial intelligence-based multipath effect detection method listed in any of the above methods.

[0147] An embodiment of the present application provides a computer program product, including a computer program that can be executed by a computer device. When the program is run on the computer device, the computer device executes an artificial intelligence-based multipath effect detection method listed in any of the above methods.

[0148] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0149] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0150] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0151] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0152] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A multi-path effect detection method based on artificial intelligence, characterized in that, Including: Determine the correlation degree between multipath error and different eigenvalues according to historical observation data; Obtain real-time observation data sent by satellites and Geographic Information System (GIS) data; Input the real-time observation data, the GIS data, and the correlation degree between the multipath error and different eigenvalues into the first model to determine the influence degree of the multipath effect on the real-time observation data; wherein, the first model is trained according to the historical observation data, the GIS data, and the correlation degree between the multipath error and different eigenvalues.

2. The method according to claim 1, wherein The determining the correlation degree between the multipath error and different eigenvalues according to the historical observation data includes: Determine the observation residual and multiple eigenvalues according to the historical observation data; wherein, the observation residual is used to indicate the multipath error; the observation residual includes single-difference pseudorange and phase residual between different stations; the eigenvalues include one or more of the following: satellite elevation angle, satellite azimuth angle, inter-frequency differential carrier-to-noise ratio, number of observed satellites, phase consistency, pseudorange consistency, pseudorange rate consistency, inter-frequency difference value of phase, or time single-difference ambiguity; Determine the correlation degree between each eigenvalue and the observation residual.

3. The method according to claim 1, characterized in that, After determining the influence degree of the multipath error on the real-time observation data, it further includes: Store the real-time observation data in the first database and update the first database.

4. The method according to claim 3, wherein After storing the real-time observation data in the first database, it further includes: Periodically train the first model according to the updated first database.

5. The method according to claim 3, wherein After storing the real-time observation data in the first database, it further includes: Periodically train the correlation model according to the updated first database; the correlation model is used to determine the correlation degree between the multipath error and different eigenvalues; The determining the correlation degree between the multipath error and different eigenvalues according to the historical observation data includes: Input the historical observation data into the correlation model to determine the correlation degree between the multipath error and different eigenvalues.

6. The method according to claim 1, characterized in that The method further includes: Obtain the observation data within the first time period sent by satellites and the GIS data within the first time period; Input the observation data within the first time period and the GIS data within the first time period into the second model to predict the influence degree of the multipath effect after the first time period.

7. The method according to claim 6, wherein The training method of the second model includes the following process: Divide the historical observation data into multiple pieces of data to be processed in the first time period according to the second time period; Determine multiple signal eigenvalues from the multiple pieces of data to be processed; wherein, each signal eigenvalue is used to indicate the influence degree of the multipath error on the corresponding data to be processed; Determine the association relationship between each signal eigenvalue and the environmental information, where the environmental information includes time information and space information; the time information is used to indicate the periodic change rule of the influence degree of the multipath error on the data to be processed; the space information is used to indicate the influence of the GIS data on the data to be processed; Train the second model according to the association relationship between each signal eigenvalue and the environmental information.

8. The method according to claim 2, wherein Determining the observation value residual according to the historical observation data includes: Obtaining the double-difference residual of the historical observation data through the double-difference positioning mode of inter-station single difference and inter-satellite single difference according to the historical observation data; wherein, the inter-station single difference refers to the differential processing of the observation data between the mobile station and the reference station; the inter-satellite single difference refers to the differential processing of the observation data between different satellites; the double-difference residual refers to the residual after the inter-station single difference and the inter-satellite single difference; Determining the observation value residual according to the double-difference residual; wherein, the observation value residual refers to the single-difference pseudorange and phase residual between the same inter-satellite and different inter-stations.

9. The method according to claim 2, wherein The method for determining the phase consistency includes the following process: Respectively obtaining the phases of the first epoch and the second epoch from the historical observation data; wherein, the first epoch is the previous epoch of the second epoch; Predicting the phase of the second epoch according to the phase of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies; Determining the phase consistency of the second epoch according to the phase of the second epoch and the predicted phase of the second epoch; wherein, the phase consistency is inversely correlated with the influence degree of the multipath error; The method for determining the pseudorange consistency includes the following process: Respectively obtaining the pseudoranges of the first epoch and the second epoch from the historical observation data; Predicting the pseudorange of the second epoch according to the pseudorange of the first epoch, the Doppler observation value of the first epoch, the Doppler observation value of the second epoch, and the carrier wavelengths of different frequencies; Determining the pseudorange consistency of the second epoch according to the pseudorange of the second epoch and the predicted pseudorange of the second epoch; wherein, the pseudorange consistency is inversely correlated with the influence degree of the multipath error.

10. The method according to claim 2, wherein The method for determining the pseudorange rate consistency includes the following process: Determining the first pseudorange rate between the first epoch and the second epoch from the historical observation data; The second pseudorange rate obtained according to the Doppler observation value; Determining the pseudorange rate consistency according to the first pseudorange rate and the second pseudorange rate; wherein, the pseudorange rate consistency is inversely correlated with the influence degree of the multipath error.

Images