A GNSS multipath error modeling method and system for a base station network

Abstract

The present invention proposes a GNSS multipath error modeling method and system for a reference station network. The present invention constructs an initial reference network; calculates a multipath error correction model for each station in the initial base station network in combination with a semi-celestial grid point model, and stores the model in a reference information library; constructs an observation subnet for this iteration; performs multipath error modeling of the semi-celestial grid point model on all stations in the observation subnet for this iteration to obtain corresponding multipath error correction models, and updates the multipath error correction model of each station in the observation subnet for this iteration to the information library; repeatedly executes the multipath error correction model construction of the adjacent station table until the multipath error modeling of each station in the observation station network is completed; and provides the information library results to the user, according to the model usage scheme of the semi-celestial grid point model method. The present invention realizes the rapid modeling of the multipath error effect of a large-scale GNSS reference station network, effectively improving the data processing service performance of the existing GNSS observation network.

Description

Technical Field

[0001] The present invention belongs to the field of global satellite navigation systems, and in particular relates to a GNSS multipath error modeling method and system for a reference station network. Background Art

[0002] Multipath error at a measuring station is a deviation in observed values ​​caused by GNSS (Global Navigation Satellite System) satellite signals reaching the receiver antenna via multiple paths due to scattering and reflection. This error is highly dependent on the satellite signal, the signal receiving equipment, and the measuring station's observation environment, making it difficult to eliminate or mitigate using differential processing, existing correction models, or parameterized estimation methods. The impact of multipath error in the observation environment can be approximated using the following simplified model. The theoretical multipath error value is related to four parameters: the distance d from the antenna phase center to the reflecting plane; the reflection coefficient α; the wavelength λ of the carrier signal; and the incident angle β of the direct signal. For carrier phase observations, the multipath error can reach up to 1 / 4 of the wavelength.

[0003] In order to model the influence of multipath errors on GNSS observation signals, the most widely used and effective method is the sidereal day filtering method. This method is based on the periodic repeatability of the geometric structure between the satellite, the station antenna and the interference source, and performs multipath error correction in the observation domain or the coordinate domain. However, since the differences in the repetition periods of different satellite orbits need to be taken into account, it is not suitable for the design of the Beidou Satellite Navigation System (BDS) composed of multiple types of high-medium-low orbit satellites. Describing the carrier phase multipath error effect at the station in the spatial domain is also an effective solution to weaken its influence. This type of modeling method reduces the influence of the differences in the repetition period of satellite orbits and is a current research hotspot. Moore et al. established an empirical model ESM (Empirical Site Model) by obtaining the residuals of the non-differenced phase observations of each station to weaken the unmodeled station-related errors. [1] Dong et al. proposed a multi-antenna shared clock method to establish a multipath error hemispherical model MHM (Multipath Hemispherical Map) [2] Zhang et al. systematically analyzed the application service solutions of MHM model in BeiDou system. [3] Since the zero mean assumption of the ESM method in obtaining the non-difference residual is unreasonable in some cases [4] Tangental. Multi-point Hemispherical Grid Model (MHGM) is a semi-spherical model of multipath error using the residuals of double-difference observations between stations. [5]The effectiveness of this method was verified using 18 years of GPS observation data from the IGS station. Compared with the aforementioned multipath error modeling methods based on undifferenced observation residuals, the MHGM method has the highest multipath error modeling accuracy and reliability.

[0004] However, the MHGM method sets the parameters to be estimated for all grid points in the semi-celestial space domain at each station. [6-7] Due to the large number of parameters and the need for global least squares estimation of all parameters to be estimated at all stations, the computational time and memory usage increase exponentially as the number of stations involved in the modeling increases. The MHGM method is therefore unable to meet the short-term and rapid update requirements of multipath error modeling for large-scale reference station networks.

[0005] Taking a 2-degree grid division scheme as an example, each station has 7,381 grid point parameters to be estimated. If the grid point parameters are solved for the entire network using double-precision floating-point data and a least-squares estimation method, the memory usage of the corresponding normal equations will increase exponentially with the number of stations involved in multipath error modeling. When the number of stations exceeds 20, the memory usage reaches over 160Gb, and when the number reaches 200, the memory usage reaches 16,200Gb. Even without considering the time-consuming problem of extremely large-scale matrix inversion, the existing MHGM method is no longer feasible for comprehensive modeling in large-scale reference station networks.

[0006] References

[0007] [1]MooreM,WatsonC,KingM,McCluskyS,TregoningP.Empiricalmodellingofsite-specificerrorsincontinuousGPSdata[J].JournalofGeodesy,2014,88(9):887-900.

[0008] [2] Dong D, Wang M, Chen W, Zeng Z, Song L, Zhang Q, Cai M, Cheng Y, Lv J.

[0009] Mitigation of multipath effect in GNSS short baseline positioning by the multipathhemispherical map [J]. Journal of Geodesy, 2016, 90 (3): 255-262.

[0010] [3] ZhangZ,DongY,WenY,LuoY.Modeling,refinementandevaluationofmultipathmitigationbasedonthehemisphericalmapinBDS2 / BDS3relativeprecisepositioning[J].Measurement, 2023.

[0011] [4]ShiQ,DaiW,ZengF,KuangC.TheBDSmultipathhemisphericalmapbasedondoubledifferenceresidualsanditsapplicationanalysis[C] / / ChinaSatelliteNavigationConference(CSNC)2016Proceedings:VolumeI.Springer,Singapore,2016:381-395.

[0012] [5]TangW,WangY,ZouX,LiY,DengC,CuiJ.VisualizationofGNSSmultipatheffectsanditspotentialapplicationinIGSdataprocessing[J].JournalofGeodesy.2021,95(9):103.

[0013] [6] Wang Yawei, Zou Xuan, Tang Weiming, Cui Jianhui, Li Yangyang. Semi-spherical grid point modeling method to reduce GNSS multipath effect [J]. Acta Geodaetica et Cartographica Sinica, 2020, 49(4): 461-468.

[0014] [7]ZouX,FuR,TangJ,WangY,FanX,LiZ,DengC,LiY,TangW.MultipatherrormitigationmethodconsideringNLOSsignalforhigh-precisionGNSSdataprocessing[J].GPSSolutions,2023,27(4).DOI:10.1007 / s10291-023-01498-2. Summary of the Invention

[0015] In order to solve the problem that the existing MHGM method is not applicable to multipath error modeling in a large-scale reference station network, the present invention proposes a GNSS multipath error modeling method and system for a reference station network.

[0016] The technical solution of the method of the present invention is a GNSS multipath error modeling method for a reference station network, and the specific steps are as follows:

[0017] Step 1: Construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network;

[0018] Step 2: Calculate the multipath error correction model at each station in the initial base station network using the semi-celestial grid point model. This model is then incorporated into the reference information library used for subsequent multipath error model calculations.

[0019] Step 3: For the reference station network consisting of the stations included in the model reference information database, the m peripheral stations are used as public stations. No more than nm stations around the public stations that do not have a multipath error correction model are selected to form the observation subnet of this iteration.

[0020] Step 4: Perform multipath error modeling using the semi-celestial grid point model for all stations in the observation subnet of this iteration to obtain the multipath error correction model for each station in the observation subnet of this iteration. Update the multipath error correction model for each station in the observation subnet of this iteration to the multipath error model reference information database.

[0021] Step 5: Repeat steps 3 and 4 until the list of m adjacent stations is traversed;

[0022] Step 6: Repeat step 5 until the multipath error modeling of each observation station in the observation station network is completed;

[0023] Step 7: Provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in the large-scale station network according to the model usage plan of the semi-celestial grid point model method.

[0024] Preferably, the planar position relationship between the observation stations is constructed in step 1 as follows:

[0025] The spatial rectangular coordinates of each observation station in the observation station network are projected onto the plane coordinate system according to the Gauss-Krüger method to form the plane position relationship between the observation stations;

[0026] Construct the initial reference network as described in step 1, as follows:

[0027] Select n observation stations in the observation station network as initial observation stations to build an initial reference network;

[0028] Among them, there is no other unselected measuring station in the network formed by n measuring stations, and the distance between stations is less than the specified threshold k;

[0029] Preferably, the calculation in step 2 is performed in combination with a semi-celestial grid point model, specifically as follows:

[0030] The n initial stations are calculated using the three-step method of inter-reference station ambiguity fixation to obtain the GNSS carrier phase double-difference observation residuals corresponding to the ambiguity fixation solutions between any two initial stations, which are further used as the input information for the calculation of the semi-celestial grid point model MHGM.

[0031] Preferably, the public site in step 3 is defined as:

[0032] The newly selected reference station subnet is a common site between the multipath model reference information database and the reference station where multipath correction model information exists;

[0033] Step 3 forms the observation subnet for this iteration. The specific steps are as follows:

[0034] Step 3.1: For the station network formed by the stations included in the model benchmark information database, obtain the network-type peripheral stations of the station network formed by the stations included in the model benchmark information database according to the plane position relationship, and form a list of adjacent stations;

[0035] Step 3.2: Select m adjacent stations as common stations, where m <n;

[0036] Step 3.3: From the observation station network for which a multipath error model has not been established, that is, from the stations not in the model reference information database, select no more than nm stations whose inter-station distance to any public site is shortest and whose maximum distance does not exceed the threshold k, to form the observation subnetwork for this iteration;

[0037] Preferably, in step 4, all observation stations in the observation subnet of this iteration are subjected to multipath error modeling using a semi-celestial grid point model, specifically as follows:

[0038] Step 4.1: Use the three-step method of inter-reference station ambiguity fixation to calculate the GNSS carrier phase double-difference observation residual corresponding to the ambiguity fixation solution between any two stations, and use it as the first type of input information for multipath modeling;

[0039] Step 4.2: Use the model parameter values ​​of the m public sites in the subnet in step 3.2 in the model benchmark information database as strong constraints for the parameters to be estimated in the multipath modeling, i.e., the second type of input information for the multipath modeling.

[0040] Step 4.3: Using the two types of input information from steps 4.1 and 4.2, calculate according to the semi-celestial grid model (MHGM) to obtain the multipath error correction model for each station in the subnet.

[0041] The technical solution of the system of the present invention is a GNSS multipath error modeling system for a reference station network, comprising:

[0042] The initial reference network construction module is used to construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network;

[0043] A multipath error correction model calculation module is used to calculate the multipath error correction model at each station in the initial base station network in combination with the semi-celestial grid point model, and to include it in the reference information library used for subsequent multipath error model calculations;

[0044] The observation subnet construction module is used to select m peripheral stations of the reference station network composed of the stations included in the model reference information database as public stations, and select no more than nm stations around the public stations that do not have a multipath error correction model established to form the observation subnet of this iteration;

[0045] An iterative observation subnet multipath error correction model construction module is used to perform multipath error modeling of the semi-celestial grid point model for all observation stations in the observation subnet of this iteration, obtain the multipath error correction model of each observation station in the observation subnet of this iteration, and update the multipath error correction model of each observation station in the observation subnet of this iteration to the multipath error model reference information library;

[0046] The station list iteration execution module repeatedly calls the observation subnet construction module and the iterative observation subnet multipath error correction model construction module until the adjacent m station lists are traversed;

[0047] The observation station network iteration execution module repeatedly calls the observation station list iteration execution module until the multipath error modeling of each observation station in the observation station network is completed;

[0048] The multipath error model benchmark information library application module is used to provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in a large-scale station network according to the model usage plan of the semi-celestial grid point model method.

[0049] The present invention decomposes the problem of GNSS multipath modeling for a large-scale reference station network into multiple smaller subnets for modeling. The multipath error models of different subnets are then spliced ​​together by setting up one or more public sites between the subnets. This significantly reduces the consumption of computing resources such as memory, CPU, and time required for modeling the entire multipath error network using the existing semi-celestial grid point model, and ensures the consistency of the multipath error modeling results between different subnets.

[0050] Compared with the existing technical means, the advantages and technical effects of the present invention are:

[0051] When modeling multipath errors using the existing semi-celestial grid point model, the number of parameters to be estimated increases significantly due to the fineness of the grid division and the number of stations involved in modeling the entire network. This also causes the memory usage of the corresponding normal equation matrix to grow exponentially, making it difficult to meet the application requirements of multipath error modeling for large-scale reference station networks. The present invention dynamically adjusts the number of stations involved in modeling for each subnet based on the performance of the data processing server. This allows for rapid and effective modeling of multipath errors for large-scale reference station networks using limited computing resources.

[0052] When performing multipath error modeling, the present invention decomposes a larger-scale reference station network into multiple smaller-scale reference station networks for separate modeling. In the modeling process, only the double-difference observation residuals of relatively short-distance baselines are involved. Since the distance between the measuring stations is too long during the baseline solution process, the double-difference ambiguity fixation result may be erroneous, and erroneous double-difference observation residuals will be obtained at this time. In addition to containing multipath errors, the carrier phase double-difference observation residual information also contains erroneous information about the ambiguity fixation result. The presence of erroneous double-difference observation residuals will affect the validity of the MHGM multipath error modeling results. Therefore, in step 3.3 of the present invention, only a smaller-scale subnet is used for multipath error modeling, which theoretically results in a higher model accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 : A flow chart of a method according to an embodiment of the present invention;

[0054] Figure 2 : A diagram of an embodiment of the present invention showing a multipath error partitioning MHGM modeling and overall splicing technology solution based on a common site;

[0055] Figure 3 : A diagram showing the improvement effect of multipath error modeling on the accuracy of Beidou system carrier phase observations according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] In specific implementation, the method proposed in the technical solution of the present invention can be automatically run by those skilled in the art using computer software technology. System devices that implement the method, such as computer-readable storage media that store the corresponding computer program of the technical solution of the present invention and computer equipment that runs the corresponding computer program, should also be within the scope of protection of the present invention.

[0058] To verify the effectiveness of the present invention, observation data of 221 base stations of the BeiDou Ground-Based Augmentation System in Guangdong Province with a sampling rate of 1 second were collected from 128 to 140 days in 2023. The post-processing version of PowerNetwork, a GNSS network RTK system management and positioning service software developed by the Satellite Navigation and Positioning Technology Research Center of Wuhan University, was used to obtain the residuals of the BeiDou system carrier phase double-difference observations between stations by fixing the station coordinates to their known true values. Taking into account that the orbital repetition period of GEO and IGSO satellites in the BeiDou system is 1 day, while the orbital repetition period of MEO satellites is 7 days, the multipath error modeling of the BeiDou system carrier phase ionospheric-free combined observations was performed using observation data from 128 to 134 days in 2023.

[0059] The following combination Figure 1-3 The specific embodiments of the present invention are introduced.

[0060] like Figure 1 As shown, the embodiment of the method of the present invention is specifically a GNSS multipath error modeling method for a reference station network, and the specific steps are as follows:

[0061] Step 1: Construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network;

[0062] Step 1 constructs the planar position relationship between observation stations as follows:

[0063] The spatial rectangular coordinates of each observation station in the observation station network are projected onto the plane coordinate system according to the Gauss-Krüger method to form the plane position relationship between the observation stations;

[0064] Construct the initial reference network as described in step 1, as follows:

[0065] Select n observation stations in the observation station network as initial observation stations to construct the initial reference network, such as Figure 2 shown.

[0066] Among them, there is no other unselected measuring station in the network formed by n measuring stations, and the distance between stations is less than the specified threshold k;

[0067] Step 1.3: The number of selected measurement stations n is set according to the server performance during data processing, and can generally be set to 4 to 10. In this embodiment, it is set to 8;

[0068] Step 1.4: Set the distance threshold k based on the latitude information of the observation station network to ensure the reliability of the inter-station ambiguity fixation results in the subsequent step 2. The threshold k should generally be less than 200 km. It can be appropriately increased in high latitudes and appropriately reduced in low latitudes. In this embodiment, it is set to 150 km.

[0069] Step 2: Calculate the multipath error correction model at each station in the initial base station network using the semi-celestial grid point model. This model is then incorporated into the reference information library used for subsequent multipath error model calculations.

[0070] The calculation described in step 2 is performed in combination with the semi-celestial grid point model, as follows:

[0071] The n initial stations are calculated using the three-step method of inter-reference station ambiguity fixation to obtain the GNSS carrier phase double-difference observation residuals corresponding to the ambiguity fixation solutions between any two initial stations, which are further used as the input information for the calculation of the semi-celestial grid point model MHGM.

[0072] Step 3: For the reference station network consisting of the stations included in the model reference information database, the m peripheral stations are used as public stations. No more than nm stations around the public stations that do not have a multipath error correction model are selected to form the observation subnet of this iteration.

[0073] The public site described in step 3 is defined as:

[0074] The newly selected reference station subnet is a common site between the multipath model reference information database and the reference station where multipath correction model information exists;

[0075] Step 3 forms the observation subnet for this iteration. The specific steps are as follows:

[0076] Step 3.1: For the station network formed by the stations included in the model benchmark information database, obtain the network-type peripheral stations of the station network formed by the stations included in the model benchmark information database according to the plane position relationship, and form a list of adjacent stations;

[0077] Step 3.2: Select m adjacent stations as common stations, where m <n;

[0078] m is set to 1 or 2, and in this embodiment is set to 2;

[0079] Step 3.3: From the observation station network for which a multipath error model has not been established, that is, from the stations not in the model reference information database, select no more than nm stations whose inter-station distance to any public site is shortest and whose maximum distance does not exceed the threshold k, to form the observation subnetwork for this iteration;

[0080] Step 4: Perform multipath error modeling using the semi-celestial grid point model for all stations in the observation subnet of this iteration to obtain the multipath error correction model for each station in the observation subnet of this iteration. Update the multipath error correction model for each station in the observation subnet of this iteration to the multipath error model reference information database.

[0081] In step 4, all observation stations in the observation subnet of this iteration are modeled with the multipath error of the semi-celestial grid point model, as follows:

[0082] Step 4.1: Use the three-step method of inter-reference station ambiguity fixation to calculate the GNSS carrier phase double-difference observation residual corresponding to the ambiguity fixation solution between any two stations, and use it as the first type of input information for multipath modeling;

[0083] Step 4.2: Use the model parameter values ​​of the m public sites in the subnet in step 3.2 in the model benchmark information database as strong constraints for the parameters to be estimated in the multipath modeling, i.e., the second type of input information for the multipath modeling.

[0084] Step 4.3: Using the two types of input information from steps 4.1 and 4.2, calculate according to the semi-celestial grid model (MHGM) to obtain the multipath error correction model for each station in the subnet.

[0085] Step 5: Repeat steps 3 and 4 until the list of m adjacent stations is traversed;

[0086] Step 6: Repeat step 5 until the multipath error modeling of each observation station in the observation station network is completed;

[0087] Since m public sites are set up between subnets to transmit model information, the consistency of the multipath error MHGM modeling results between different subnets can be guaranteed;

[0088] Step 7: Provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in the large-scale station network according to the model usage plan of the semi-celestial grid point model method.

[0089] The technical solution of the system of the present invention is a GNSS multipath error modeling system for a reference station network, comprising:

[0090] The initial reference network construction module is used to construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network;

[0091] A multipath error correction model calculation module is used to calculate the multipath error correction model at each station in the initial base station network in combination with the semi-celestial grid point model, and to include it in the reference information library used for subsequent multipath error model calculations;

[0092] The observation subnet construction module is used to select m peripheral stations of the reference station network composed of the stations included in the model reference information database as public stations, and select no more than nm stations around the public stations that do not have a multipath error correction model established to form the observation subnet of this iteration;

[0093] An iterative observation subnet multipath error correction model construction module is used to perform multipath error modeling of the semi-celestial grid point model for all observation stations in the observation subnet of this iteration, obtain the multipath error correction model of each observation station in the observation subnet of this iteration, and update the multipath error correction model of each observation station in the observation subnet of this iteration to the multipath error model reference information library;

[0094] The station list iteration execution module repeatedly calls the observation subnet construction module and the iterative observation subnet multipath error correction model construction module until the adjacent m station lists are traversed;

[0095] The observation station network iteration execution module repeatedly calls the observation station list iteration execution module until the multipath error modeling of each observation station in the observation station network is completed;

[0096] The multipath error model benchmark information library application module is used to provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in a large-scale station network according to the model usage plan of the semi-celestial grid point model method.

[0097] The initial reference network construction module, multipath error correction model calculation module, observation subnet construction module, iterative observation subnet multipath error correction model construction module, observation station list iterative execution module, observation station network iterative execution module, and multipath error model reference information library application module are all deployed on the server;

[0098] Considering server performance during multipath modeling, a total of 33 subnets were created, with a maximum number of sites n per subnet of 9. The multipath error model was spliced ​​by setting up two common sites between different subnets to form the multipath error correction model for the Guangdong Province BeiDou Ground-Based Augmentation System.

[0099] PowerNetwork software was used to reprocess observation data from days 128 to 140 in 2023. The BeiDou MHGM multipath error correction model at each reference station was used during data processing to correct the residuals of the inter-reference station carrier phase double-difference observations. The effectiveness of the multipath model correction results was statistically analyzed. Data from days 128 to 134 represent modeling days, while data from days 135 to 140 represent extrapolated test results.

[0100] from Figure 3 Statistics from the 13-day test results show that after implementing multipath error model corrections, the Beidou system's carrier phase observation residuals achieved a certain degree of accuracy improvement. The RMS on the modeling day decreased from an average of 2.0 cm to 1.5 cm, with an average improvement of 25.0%. During the subsequent six days of extrapolation testing, the RMS residuals of the Beidou system's carrier phase double-difference observations decreased from an average of 2.2 cm to 1.9 cm, achieving an improvement of 11.2%.

[0101] Therefore, according to this patent, it is feasible to model BeiDou system multipath errors across a large-scale base station network. Without installing additional hardware, the performance of the existing Guangdong Province BeiDou ground-based augmentation system can be improved by 11.2%. This has significant scientific research and engineering application value for improving the accuracy and reliability of existing BeiDou / GNSS system data processing services.

[0102] It should be understood that parts not elaborated in detail in this specification belong to the prior art.

[0103] It should be understood that the above description of the embodiments is relatively detailed and cannot be regarded as limiting the scope of protection of the patent of the present invention. Under the guidance of the present invention, ordinary technicians in this field can also make substitutions or modifications without departing from the scope of protection of the claims of the present invention, which all fall within the scope of protection of the present invention. The scope of protection requested by the present invention shall be based on the attached claims.

Claims

1. A GNSS multipath error modeling method for a reference station network, characterized in that: The following steps are involved: Step 1: Construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network; Step 2: Calculate the multipath error correction model at each station in the initial base station network using the semi-celestial grid point model. This model is then incorporated into the reference information library used for subsequent multipath error model calculations. Step 3: For the reference station network consisting of the stations included in the model reference information database, the m peripheral stations are used as public stations. No more than nm stations around the public stations that do not have a multipath error correction model are selected to form the observation subnet of this iteration. Step 4: Perform multipath error modeling using the semi-celestial grid point model for all stations in the observation subnet of this iteration to obtain the multipath error correction model for each station in the observation subnet of this iteration. Update the multipath error correction model for each station in the observation subnet of this iteration to the multipath error model reference information database. Step 5: Repeat steps 3 and 4 until the list of m adjacent stations is traversed; Step 6: Repeat step 5 until the multipath error modeling of each observation station in the observation station network is completed; Step 7: Provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in the large-scale station network according to the model usage plan of the semi-celestial grid point model method.

2. The GNSS multipath error modeling method for a reference station network according to claim 1, wherein: Its characteristics are: Step 1 constructs the planar position relationship between observation stations as follows: The spatial rectangular coordinates of each observation station in the observation network are projected into the plane coordinate system according to the Gauss-Krüger method to form the plane position relationship between the observation stations.

3. The GNSS multipath error modeling method for a reference station network according to claim 2, wherein: Its characteristics are: Construct the initial reference network as described in step 1, as follows: Select n observation stations in the observation station network as initial observation stations to build an initial reference network; In this case, there should be no other unselected measuring stations in the network formed by the n measuring stations, and the distances between the stations should be less than the specified threshold k.

4. The GNSS multipath error modeling method for a reference station network according to claim 3, wherein: Its characteristics are: The calculation described in step 2 is performed in combination with the semi-celestial grid point model, as follows: The n initial stations are calculated using the three-step method of ambiguity fixation between reference stations to obtain the GNSS carrier phase double-difference observation residuals corresponding to the ambiguity fixation solutions between any two initial stations, which are further used as the input information for the semi-celestial grid point model MHGM calculation.

5. The GNSS multipath error modeling method for a reference station network according to claim 4, wherein: Its characteristics are: The public site described in step 3 is defined as: The newly selected reference station subnet is a common site between the multipath model reference information library and the reference station with multipath correction model information.

6. The GNSS multipath error modeling method for a reference station network according to claim 5, characterized in that: Its characteristics are: Step 3 forms the observation subnet for this iteration. The specific steps are as follows: Step 3.1: For the station network formed by the stations included in the model benchmark information database, obtain the network-type peripheral stations of the station network formed by the stations included in the model benchmark information database according to the plane position relationship, and form a list of adjacent stations; Step 3.2: Select m adjacent stations as common stations, where m <n; Step 3.3: From the observation station network for which a multipath error model has not been established, that is, from the stations not in the model reference information database, select no more than nm stations whose inter-station distance to any public site is the shortest and whose maximum distance does not exceed the threshold k to form the observation subnetwork for this iteration.

7. The GNSS multipath error modeling method for a reference station network according to claim 6, wherein: Its characteristics are: In step 4, all observation stations in the observation subnet of this iteration are modeled with the multipath error of the semi-celestial grid point model, as follows: Step 4.1: Use the three-step method of inter-reference station ambiguity fixation to calculate the GNSS carrier phase double-difference observation residual corresponding to the ambiguity fixation solution between any two stations, and use it as the first type of input information for multipath modeling; Step 4.2: Use the model parameter values ​​of the m public sites in the subnet in step 3.2 in the model benchmark information database as the strong constraint information for the parameters to be estimated in the multipath modeling, i.e., the second type of input information for the multipath modeling. Step 4.3: Using the two types of input information from steps 4.1 and 4.2, calculate according to the semi-celestial grid point model to obtain the multipath error correction model for each station in the subnet.

8. A GNSS multipath error modeling system for a reference station network, characterized in that: include: The initial reference network construction module is used to construct the plane position relationship between the observation stations, and select n stations in the observation station network as n initial stations to construct the initial reference network; A multipath error correction model calculation module is used to calculate the multipath error correction model at each station in the initial base station network in combination with the semi-celestial grid point model, and to include it in the reference information library used for subsequent multipath error model calculations; The observation subnet construction module is used to select m peripheral stations of the reference station network composed of the stations included in the model reference information database as public stations, and select no more than nm stations around the public stations that do not have a multipath error correction model established to form the observation subnet of this iteration; An iterative observation subnet multipath error correction model construction module is used to perform multipath error modeling of the semi-celestial grid point model for all observation stations in the observation subnet of this iteration, obtain the multipath error correction model of each observation station in the observation subnet of this iteration, and update the multipath error correction model of each observation station in the observation subnet of this iteration to the multipath error model reference information library; The station list iteration execution module repeatedly calls the observation subnet construction module and the iterative observation subnet multipath error correction model construction module until the adjacent m station lists are traversed; The observation station network iteration execution module repeatedly calls the observation station list iteration execution module until the multipath error modeling of each observation station in the observation station network is completed; The multipath error model benchmark information library application module is used to provide the multipath error model benchmark information library results to users, and provide accurate and effective multipath error model correction information for all stations in a large-scale station network according to the model usage plan of the semi-celestial grid point model method.

Images