RTK positioning method based on the fusion of geometric model algorithm and non-geometric model algorithm

By combining geometric model and geometric modelless algorithm, we establish a relational expression between satellite and user position parameters, and select the best high-quality satellite for RTK positioning, solving the problem of the impact of satellite observation quality in multimode and multi-frequency environments, achieving high-precision and High reliability GNSS positioning.

CN114779303BActive Publication Date: 2024-11-08THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210256583.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-16
Publication Date
2024-11-08
Estimated Expiration
2042-03-16

AI Technical Summary

Technical Problem

When the existing GNSS satellite RTK positioning algorithm is used to deal with multimode and multifrequency conditions, it is difficult to effectively identify and eliminate poor-quality satellite observation information, affecting positioning accuracy and reliability.

Method used

Using a fusion method based on geometric model algorithm and geometric model-free algorithm, the user location and satellite state parameters are processed through Kalman filtering modeling, and the relationship expression of the single difference between satellites and the user baseline parameters is established. , choose the best high-quality satellite for positioning and solving.

Benefits of technology

It realizes high-precision modeling and high-reliability positioning of user position parameters in a multi-mode and multi-frequency GNSS satellite environment, avoids the impact of poor quality satellites on position parameters, and adapts to the development trend of navigation satellites with multi-mode and multi-frequency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114779303B_ABST
    Figure CN114779303B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of GNSS satellite RTK positioning, and specifically to an RTK positioning method based on the fusion of a geometric model algorithm and a non-geometric model algorithm. The implementation process is: perform Kalman filter modeling on the user's position parameters to establish a geometric model GB; perform Kalman filter modeling on the satellite's state parameters to establish a non-geometric model GF; filter and update the satellite state parameters in the non-geometric model GF according to the satellite carrier and pseudo-range observation values; establish the connection between the position parameters of the geometric model GB and the satellite state parameters of the non-geometric model GF; perform filtering update or least squares solution to obtain the user's position parameters. The present invention can avoid the influence of satellites with poor quality on the solution of position parameters, timely eliminate satellites with poor observation quality, and will not affect the geometric model of the user's state parameters.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of GNSS satellite RTK positioning, specifically to an RTK positioning method based on the fusion of geometric model algorithms and geometric model-free algorithms. Background Technology

[0002] Currently, the number of GNSS satellites and frequencies is increasing daily, allowing terminals to utilize more navigation satellite information for RTK positioning. This abundance of redundant observation information makes it possible for users to obtain high-precision and highly reliable positioning results. However, the terminal's computing power is limited and cannot, and does not need to, process all observation information. Therefore, it is necessary to differentiate the observation quality of different satellites and select the best ones for user position calculation.

[0003] Algorithms for RTK positioning using GNSS satellites are divided into geometric model algorithms (GB) and geometric model-less algorithms (GF). The geometric model algorithm directly uses user coordinates as filtering parameters, but the double-difference observations from GNSS satellites are highly coupled, making it impossible to avoid the influence of lower-quality satellites on the position parameter calculation. The geometric model-less algorithm uses satellite state as filtering parameters, allowing for the selection of higher-quality satellites for calculation, but it does not model the user coordinate parameters.

[0004] Based on the above reasons, an RTK positioning algorithm based on the fusion of geometric model algorithm and geometric model-less algorithm is proposed. This algorithm can both model user coordinate parameters using geometric model algorithm and select high-quality satellite parameters for solution using geometric model-less algorithm, thus effectively handling the situation of multi-mode and multi-frequency GNSS satellites. Summary of the Invention

[0005] The purpose of this invention is to provide an RTK positioning method based on the fusion of geometric model algorithms and non-geometric model algorithms, thereby effectively handling the situation of multi-mode and multi-frequency GNSS satellites.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] An RTK localization method based on the fusion of geometric model algorithms and geometric model-free algorithms includes the following steps:

[0008] Step 1: Perform Kalman filtering modeling on the user's location parameters to establish the geometric model GB;

[0009] Step 2: Perform Kalman filter modeling on the satellite's state parameters to establish a geometry-free model (GF).

[0010] Step 3: Filter and update the satellite state parameters in the geometrically undefined model GF based on satellite carrier and pseudorange observations;

[0011] Step 4: Establish the relationship between the position parameters of the geometric model GB and the satellite state parameters without the geometric model GF, and obtain the relationship expression between the inter-station single-difference satellite-to-ground distance parameters and the user baseline parameters;

[0012] Step 5: Select satellite state parameters with observation quality higher than the threshold as the observed values ​​of user position parameters in the geometric model GB, and filter and update or solve the expression obtained in Step 4 to obtain the user's position parameters.

[0013] Furthermore, the specific implementation process of step four is as follows:

[0014] The user baseline parameters that need to be finally solved are:

[0015] X = [ΔX, ΔY, ΔZ, ΔT] ru,w ]

[0016] Where ΔX, ΔY, and ΔZ represent the user baseline coordinate components, and ΔT represents the user baseline coordinate components, respectively. ru,w Indicates the tropospheric moisture component of the single difference between stations;

[0017] The relationship between the inter-satellite station single-difference satellite-to-ground distance parameter and the user baseline parameter is expressed as follows:

[0018]

[0019] in:

[0020] This represents the double-difference satellite-to-ground distance parameter, where j represents the reference satellite, k represents other satellites, r represents the base station, and u represents the rover station. and These represent the single-difference satellite-to-ground distances between satellite k and satellite j, respectively.

[0021] and Let X represent the partial derivatives of the satellite distances from satellite j and satellite k with respect to the user baseline coordinate parameter ΔX, respectively. k Represents the three-dimensional X coordinate of satellite k, X j Represents the three-dimensional X coordinate of satellite j, X u Represents the user's three-dimensional X-coordinate;

[0022] and Let Y represent the partial derivatives of the satellite distances from satellite j and satellite k with respect to the user baseline coordinate parameter ΔY. k Represents the three-dimensional Y-coordinate of satellite k, Y j This represents the three-dimensional Y-coordinate of satellite j, where Y... u Represents the user's three-dimensional Y-coordinate;

[0023] and Let Z represent the partial derivatives of the satellite distances from satellite j and satellite k with respect to the user baseline coordinate parameter ΔZ, respectively. k Represents the three-dimensional Z-coordinate of satellite k, Z j Represents the three-dimensional Z-coordinate of satellite j, Z u Represents the user's three-dimensional Z-coordinate;

[0024] and Let represent the tropospheric wet delay mapping functions for satellite j and satellite k, respectively. Further, in step five, if the least squares method is used, the position parameters are solved according to the following formula:

[0025] X = (B T PB) -1 (B T Pl)

[0026] Where B is the coefficient matrix composed of the observation equations listed by the selected satellites, and P is the double-difference satellite-to-ground distance parameter. The inverse of the variance matrix, and the residual vector of the observation equations of the selected satellite.

[0027] The present invention has the following advantages:

[0028] 1. This invention takes into account the advantages of modeling user coordinate parameters in the traditional geometric model (GB) algorithm, and can take into account the traditional model algorithm.

[0029] 2. Based on this, the present invention further considers the advantages of the geometric model-free (GF) algorithm in modeling satellite-related state parameters. It can avoid the influence of poor-quality satellites on the position parameter solution, promptly remove satellites with poor observation quality, and will not affect the geometric model (GB) of user state parameters.

[0030] 3. This invention adapts to the development trend of multi-mode and multi-frequency GNSS navigation satellites, allowing users to freely select satellites and frequencies for calculation based on the terminal's computing power. Attached Figure Description

[0031] Figure 1 This is a flowchart of an RTK localization method based on the fusion of geometric model algorithm and non-geometric model algorithm according to an embodiment of the present invention. Detailed Implementation

[0032] The invention will be further explained below with reference to the accompanying drawings.

[0033] An RTK localization method based on the fusion of geometric model algorithms and geometric model-free algorithms, comprising the following steps:

[0034] Step 1: Perform Kalman filter modeling on the user's location parameters, mainly establishing, maintaining, and updating the geometric model (GB). The following parameters are established for the user's location information:

[0035] X GB =[ΔX, ΔY, ΔZ, V] x V y V z ΔT ru,w ] T

[0036] Where ΔX, ΔY, and ΔZ represent the user baseline coordinate components, respectively, and V x V y V z These represent the user velocity components, ΔT and ΔT, respectively. ru,w This represents the tropospheric moisture component of the single-difference tropospheric difference between stations.

[0037] The state transition equation for the user location parameters is expressed as follows:

[0038] X GB (k)=Φ GB (k, k-1)·X GB (k-1), D e,GB (k-1)

[0039] Where, Φ GB (k, k-1) is the state transition matrix, with the following specific form; D e,GB (k-1) represents the noise in the state transition process.

[0040]

[0041] Δt represents the difference between time k-1 and time k.

[0042] Step 2: Perform Kalman filter modeling for the satellite's state parameters, primarily establishing and maintaining a geometry-free (GF) model. The following parameters are established for the satellite's state information:

[0043]

[0044] Among them, X GF For all satellite state parameters, Let m be the number of satellites and their state parameters be m. The state parameters of each satellite are expressed as follows:

[0045]

[0046] Where r represents the base station, u represents the rover station, and s represents any satellite. Indicates the single-difference satellite distance between stations. Indicates the single-difference ionosphere between stations. Indicates the L1 ambiguity of the single difference between stations. Indicates the ambiguity of the single-difference width lane between stations. This indicates the ambiguity of the single-difference ultra-wide alleyway between stations.

[0047] The satellite state parameter state transition equation is expressed as:

[0048] X GF (k)=Φ GF (k, k-1)·X GF (k-1), D e,GF (k-1)

[0049] Where, Φ GF (k, k-1) is the state transition matrix, where Φ GF (k, k-1) = I; D e,GF (k-1) represents the noise in the state transition process.

[0050] Step 3: Filter and update the satellite state parameters from Step 2 based on the satellite carrier and pseudorange observations.

[0051] The carrier and pseudorange observations are expressed as follows:

[0052]

[0053] Where j represents the reference satellite, k represents other satellites, i represents the frequency number (i = 1, w, e), w represents the wide lane, and e represents the extra-wide lane. This represents the double-difference pseudorange observation at any frequency i. The observation noise for this pseudorange observation. Represents the double-difference carrier observation at any frequency i. For the observation noise of this carrier observation, λ i Represents the wavelength of an observation at any frequency i.

[0054] Step 4: Establish the relationship between the positional parameters of the geometric model (GB) in Step 1 and the state parameters of the non-geometric model (GF) in Step 2.

[0055] The user baseline parameters that need to be finally solved are:

[0056] X = [ΔX, ΔY, ΔZ, ΔT] ru,w ]

[0057] The relationship between the inter-station single-difference satellite-to-ground distance parameters of the two satellites j and k obtained in step three and the user baseline parameters is expressed as follows:

[0058]

[0059] in:

[0060] This represents the double-difference satellite-to-ground distance parameter;

[0061] and Let X and K represent the partial derivatives of the distance between satellites j and k with respect to the user baseline coordinate parameter ΔX, respectively. k Represents the three-dimensional X coordinate of satellite k, X j Represents the three-dimensional X coordinate of satellite j, X u Represents the user's three-dimensional X-coordinate;

[0062] and Let Y and Y represent the partial derivatives of the distances between satellites j and k with respect to the user baseline coordinate parameter ΔY, respectively. k Represents the three-dimensional Y-coordinate of satellite k, Y j This represents the three-dimensional Y-coordinate of satellite j, where Y... u Represents the user's three-dimensional Y-coordinate;

[0063] and Let Z represent the partial derivatives of the distances between satellites j and k with respect to the user baseline coordinate parameter ΔZ, respectively. k Represents the three-dimensional Z-coordinate of satellite k, Z j Represents the three-dimensional Z-coordinate of satellite j, Z t Represents the user's three-dimensional Z-coordinate;

[0064] and Let represent the tropospheric wet delay mapping functions for satellites j and k, respectively.

[0065] Step 5: Based on Step 3, the satellite-to-ground distance state parameters of all satellites can be obtained. However, since there are many GNSS navigation satellites, users can select the state parameters of satellites with higher observation quality as the observed values ​​of user position parameters in the geometric model (GB) of Step 1, and then filter and update them or solve them using least squares.

[0066] The observation equations listed in step four are used to update the position using Kalman filtering; if the least squares method is used, the position parameters are solved according to the following formula:

[0067] X = (B T PB) -1 (B T Pl)

[0068] Where B is the coefficient matrix composed of the observation equations listed by the selected satellites, and P is the double-difference satellite-to-ground distance parameter. The inverse of the variance matrix, and the residual vector of the observation equations of the selected satellite.

[0069] After calculating the baseline vector X, the final location coordinates of the rover user are obtained based on the coordinates of the base station.

Claims

1. An RTK positioning method based on the fusion of geometric model algorithm and non-geometric model algorithm, characterized in that: The following steps are involved: Step 1: Perform Kalman filter modeling on the user's position parameters and establish a geometric model GB; Step 2: Perform Kalman filter modeling on the state parameters of the satellite and establish a geometry-free model GF; Step 3: Filter and update the satellite state parameters in the geometry-free model GF according to the satellite carrier and pseudorange observation values; Step 4: Establish the relationship between the position parameters of the geometric model GB and the satellite state parameters without the geometric model GF, and obtain the relationship expression between the inter-station single-difference satellite-to-earth distance parameters and the user baseline parameters; Step 5: Select satellite state parameters with observation quality higher than a threshold as observation values ​​of user position parameters in the geometric model GB, perform filtering update or least squares solution on the expression obtained in step 4 to obtain the user position parameters; Among them, the specific implementation process of step 4 is: The user baseline parameters that need to be finally solved are: X=[ΔX,ΔY,ΔZ,ΔT ru,w ] Among them, ΔX, ΔY, ΔZ represent the user baseline coordinate components, ΔT ru,w represents the single-difference tropospheric wet component between stations; The relationship between the single difference satellite-to-ground distance parameter between satellite stations and the user baseline parameter is expressed as: in: represents the double-difference satellite-to-ground distance parameter, j represents the reference satellite, k represents other satellites, r represents the base station, and u represents the mobile station. and They represent the inter-station single difference satellite-to-earth distances of satellite k and satellite j respectively; and denote the partial derivatives of the satellite j and satellite k’s distance from the ground to the user’s baseline coordinate parameter ΔX, respectively. k represents the three-dimensional X coordinate of satellite k, X j represents the three-dimensional X coordinate of satellite j, X u Indicates the three-dimensional X coordinate of the user's position; and represents the partial derivative of the satellite j and satellite k's distance from the ground to the user's baseline coordinate parameter ΔY, Y k represents the three-dimensional Y coordinate of satellite k, Y j represents the three-dimensional Y coordinate of satellite j, Y u Indicates the three-dimensional Y coordinate of the user's position; and are the partial derivatives of the satellite j and satellite k’s distance from the ground with respect to the user baseline coordinate parameter ΔZ, respectively. k represents the three-dimensional Z coordinate of satellite k, Z j represents the three-dimensional Z coordinate of satellite j, Z u Indicates the three-dimensional Z coordinate of the user's position; and represent the tropospheric wet delay mapping functions of satellite j and satellite k respectively.

2. The RTK positioning method based on the fusion of geometric model algorithm and non-geometric model algorithm according to claim 1 is characterized in that: In step 5, if the least squares method is used, the location parameter is solved according to the following formula: X=(B T PB) -1 (B T Pl) Among them, B is the coefficient matrix composed of the observation equations listed for the selected satellite, and P is the double-difference satellite-to-earth distance parameter is the inverse of the variance matrix of , and l is the residual vector of the observation equation listed for the selected satellite.