CN110007320B

CN110007320B - Network RTK resolving method

Info

Publication number: CN110007320B
Application number: CN201910291020.1A
Authority: CN
Inventors: 何锡扬; 崔红正
Original assignee: Qianxun Spatial Intelligence Inc
Current assignee: Qianxun Spatial Intelligence Inc
Priority date: 2017-04-12
Filing date: 2017-04-12
Publication date: 2022-12-20
Anticipated expiration: 2037-04-12
Also published as: CN107422343B; CN110007320A; CN107422343A

Abstract

The invention discloses a network RTK resolving method, which combines a baseline resolving module and an atmosphere model module in the conventional network RTK technology for joint resolving, namely, a same filter is used for simultaneously estimating all baseline parameters and atmosphere model parameters; meanwhile, the constraint of a double-difference integer ambiguity closed loop is directly imposed in the joint solution. The method can improve the accuracy and reliability of baseline double-difference integer ambiguity fixing in the network RTK, and solve the problem that the performance of the network RTK is reduced due to low baseline ambiguity fixing rate in partial areas.

Description

Network RTK resolving method

The application is a divisional application with the application number of CN201710237878.0 and the application date of 2017, 4, 12 and named as a network RTK solution method.

Technical Field

The invention relates to a positioning technology, in particular to a network RTK resolving method, which is used for improving the accuracy and reliability of baseline integer ambiguity fixing in the network RTK and solving the problem that the performance of the network RTK is reduced due to the fact that the baseline integer ambiguity fixing rate in a partial area is not high.

Background

One of the main sources of errors in Satellite positioning in Global Navigation Satellite System (GNSS) is the extra atmospheric delay caused by atmospheric refraction when Satellite signals propagate from the satellites to the receiver, and the atmospheric delay mainly includes the ionospheric delay and the tropospheric delay caused by the ionosphere and the troposphere. Paths of satellite signals transmitted by one satellite to two receivers with similar geographic positions are basically similar, so that atmospheric delay caused by atmospheric refraction is mostly eliminated after the difference. The differential relative positioning technology can accurately determine the relative position between two adjacent receivers by using the observed value after the phase difference of the two receivers, and can accurately determine the position of the other receiver if the absolute position of one receiver is accurately known. The receivers with known positions are generally referred to as reference stations and the receivers with the desired positions are generally called rover stations, or rover users.

The Real-time kinematic (RTK) technique based on a carrier phase observation is a relative positioning technique capable of providing Real-time high precision (centimeter level), and the basic principle is a carrier phase split relative positioning technique, which mainly uses the high-precision carrier phase observation to achieve centimeter level relative positioning. The main limitation of the conventional RTK technology is that as the distance between the base station and the mobile user increases, the propagation paths cancel smaller and smaller, thereby causing the relative positioning accuracy to deteriorate, and the operation radius of the conventional RTK is generally limited to 10 km. In order to increase the operation range of the conventional RTK technology, around 2000 years, a base station network is formed by a plurality of base stations, and high-precision differential correction data is provided for a rover within the coverage area of the network, so that high-precision positioning of the rover, namely the network RTK technology, is realized. The coordinates of the GNSS network reference station are known in a coordinate frame, the atmospheric model correction value in the GNSS reference station network can be calculated through certain data and a physical model, the atmospheric model correction value is broadcasted to the rover stations in the network coverage range, and the high-precision coordinates of the rover stations in the same coordinate frame can be determined by using a difference principle.

GNSS reference stations, or GNSS reference stations, often referred to as GNSS CORS tracking stations, are generally known in coordinates (under a certain coordinate frame), have a good surrounding environment, are mainly free of physical or electromagnetic interference (e.g. trees, metals, electromagnetic jammers, etc.) around, and are equipped with high performance geodetic GNSS dual-frequency (or multi-frequency) receivers.

A network of GNSS reference stations is formed by a plurality of physically adjacent GNSS reference stations connected together. A network of GNSS reference stations typically has a minimum of 3 GNSS reference stations. The distance between the GNSS reference stations is about 70 kilometers (typical distance of a medium latitude area; the distance between stations should be closer on average in a low latitude area, about 50-60 kilometers approximately; the distance between stations at a medium latitude can be larger), a certain area (one or more administrative areas) is covered, a plurality of regional GNSS reference station networks are connected together to form a larger GNSS reference station network, so that the GNSS reference station network covers a city, one or more provinces can be expanded to the whole country or a plurality of countries, and even a global GNSS reference station network is formed.

And generating observation data of the VRS virtual reference station according to the GNSS reference station data and the atmosphere model calculated by the GNSS reference station network. The initial coordinates of the rover station are typically selected as the coordinates of the virtual reference station to ensure that the baseline distance formed by the virtual reference station and the rover station is short. And when the rover station is beyond a certain range from the current VRS station, regenerating a VRS station closer to the rover station for the user to use. After receiving VRS observation data broadcasted by a GNSS reference station network, the rover station can eliminate most GNSS errors such as atmospheric delay, satellite orbit and clock error, satellite hardware delay and the like through difference, so that the real-time high-precision coordinates of the rover station are obtained.

The baseline solution module and the atmospheric model module of the conventional network RTK solution are performed in two steps. Firstly, floating solution resolving is carried out on each base line independently, the resolved ambiguity floating solution is used for searching, and the ambiguity of each base line is fixed respectively. And after the integer ambiguity is fixed, integrating all baseline calculation results in the GNSS reference station network to model the regional atmosphere correction. The detailed steps of the solution are as follows:

1) Baseline calculation (Single baseline)

And respectively solving double-difference floating ambiguity and double-difference atmospheric (ionosphere and troposphere) delay by using double-difference carrier phase observed values for each base line in the GNSS reference station network.

2) Integer ambiguity search and fix (Single Baseline)

And searching each base line in the ambiguity candidate range respectively according to the double-difference integer ambiguity floating solution estimated in the last step, and fixing the integer ambiguity. Because the double difference integer ambiguity has strong correlation, a decorrelation method (e.g., LAMBDA) algorithm is generally adopted to perform decorrelation processing so as to narrow the search range.

3) Using fixed integer ambiguity as constraint (single base line)

And (4) re-calculating more accurate single-baseline double-difference atmosphere (ionosphere and troposphere) delay correction by taking the fixed integer ambiguity as a constraint condition.

4) Atmospheric (ionospheric and tropospheric) model solution

And comprehensively considering all baseline double-difference atmosphere (ionosphere and troposphere) delay calculation results in the GNSS reference station network, and modeling atmosphere (ionosphere and troposphere) delay in the area range by using an area gradient model. The calculated atmosphere (ionosphere and troposphere) model is used for delaying and correcting the atmosphere (ionosphere and troposphere) of the rover, and the positioning accuracy is improved.

5) Generation of VRS virtual observatory data

And selecting a GNSS base station as a reference base station, correcting observation data of the reference base station according to the VRS virtual station coordinates and the calculated atmosphere (ionosphere and troposphere) model, generating observation data of the VRS station, and broadcasting the observation data to users in real time.

The disadvantages and limitations of the prior art are mainly due to the fact that the multi-baseline closure condition is not fully utilized to optimize the fixation of the integer ambiguity during the fixation of the integer ambiguity and the atmospheric (ionosphere and troposphere) model solution. The case of the single baseline integer ambiguity fix error is difficult to be found. The use of a wrongly fixed integer ambiguity for the calculation of the atmosphere (ionosphere and troposphere) model directly reduces the positioning accuracy of the rover.

Disclosure of Invention

The invention aims to provide a network RTK resolving method, which improves the accuracy and reliability of baseline double-difference integer ambiguity fixing in network RTK and solves the problem of low baseline double-difference integer ambiguity fixing rate in partial regions.

Another object of the present invention is to provide a network RTK solution method, which improves the model accuracy of double difference atmosphere (ionosphere and troposphere) delay, thereby better and more stably ensuring the positioning accuracy of the rover.

A network RTK resolving method utilizes a closed-loop condition of double-difference integer ambiguity floating-point ambiguity of a base station network to constrain integer ambiguity floating-point ambiguity resolution, and can improve the precision of integer ambiguity floating-point ambiguity.

And in the process of fixing the integer ambiguity, reinforcing the integer ambiguity closed-loop constraint condition to select an integer ambiguity fixing solution.

And meanwhile, the accuracy of integer ambiguity fixing of all baselines is improved, so that the accuracy of atmosphere (ionosphere and troposphere) model solution is improved, and a more reliable network RTK atmosphere (ionosphere and troposphere) correction value is provided.

Adding stochastic model constraints of atmosphere (ionosphere and troposphere) delay in the baseline joint solution stage helps to improve the precision of the integer ambiguity floating solution and the atmosphere (ionosphere and troposphere) model.

Another network RTK solution method includes:

step 1: according to the scale factors of the atmosphere (ionosphere and troposphere), carrying out joint solution on the parameters of a baseline and atmosphere (ionosphere and troposphere) delay model involved in the GNSS network,

parameters (such as receiver clock error and carrier ambiguity) of various baselines formed between base stations in the network RTK and double-difference atmosphere (ionosphere and troposphere) delay model parameters are jointly solved, namely, the same filter is used for simultaneously estimating all the baseline parameters and the atmosphere delay model parameters. The initial values of the model parameters of the atmosphere (ionosphere and troposphere) are set based on historical experience. In addition, random model constraints are respectively carried out on the ionosphere delay and the troposphere delay by using the ionosphere scale factors and the troposphere scale factors;

step 2: floating ambiguity closed-loop constraints

Adding a double-difference floating ambiguity closed-loop constraint condition, namely setting the closed-difference theoretical value of the double-difference floating ambiguity of each base line in the base station network to be zero so as to improve the estimation precision of the multi-base-line floating ambiguity solution and provide better initial conditions for integer ambiguity search and fixation;

and step 3: searching for whole-cycle double-difference ambiguities

Carrying out integer ambiguity search on the floating ambiguity of each base line in the reference station network, reducing the calculation amount by using the most widely used LAMBDA algorithm to reduce the correlation between ambiguities through ambiguity reduction correlation change in order to reduce the search range and reduce the calculation amount due to strong correlation of ambiguities, and generating an integer ambiguity candidate subset according to the search range;

and 4, step 4: whole-cycle ambiguity closed-loop inspection

Performing ambiguity closed-loop inspection on all the candidates in the whole-cycle ambiguity candidate subset, wherein the ambiguity residual error is minimum, and the whole-cycle ambiguity candidate group with all satellites meeting the closed-loop inspection is selected as a final ambiguity fixed solution;

and 5: constraining integer ambiguity, and solving atmosphere (ionosphere and troposphere) delay model parameters

Solving double-difference atmosphere (troposphere and ionosphere) delay of each satellite pair of each baseline by taking the ambiguity fixed solution as a constraint condition, and establishing an atmosphere (ionosphere and troposphere) delay model by utilizing the solved double-difference atmosphere (ionosphere and troposphere) delay;

the atmospheric scale factor is solved in real time by epoch, and is fed back to the step 1 for updating, and the atmospheric scale factor is used as the atmospheric delay stochastic model constraint in the baseline joint solution of the next epoch; such as; calculating and updating the atmospheric scale factor of the double-difference atmospheric delay of each satellite pair of each solved baseline, and taking the atmospheric scale factor as step 1 to continue to serve as atmospheric delay random model constraint in the baseline joint solution of the next epoch;

step 6: virtual Reference Stations (VRS) station observation data generation

And selecting a GNSS reference station as a reference base station, correcting observation data of the reference base station according to the VRS virtual station coordinates and the calculated atmosphere (ionosphere and troposphere) delay model, generating observation data of the VRS station, and broadcasting the observation data to users in real time.

According to the technical scheme, a baseline resolving module and an atmosphere model module in the conventional network RTK technology are combined together for combined resolving, namely, the same filter is used for simultaneously estimating all baseline parameters and all atmosphere model parameters; meanwhile, the constraint of double-difference integer ambiguity closed loops is directly imposed during the joint solution; random model constraint of the double-difference atmosphere delay model is added; and the atmospheric scale factor is solved in real time by adopting epoch by epoch, and is fed back to the next epoch. Therefore, the calculation of the double-difference integer ambiguity is more reliable, the calculated atmosphere delay model and the network RTK differential correction data can be more reliable according to the atmosphere delay model, and a mobile user can have better user experience (precision and reliability).

Drawings

Fig. 1 is a flowchart of a network RTK solution method according to the present invention.

Detailed Description

The network RTK solution method of the present invention is further described in detail below with reference to the accompanying drawings.

Fig. 1 is a flowchart of a network RTK solution method according to the present invention, and as shown in fig. 1, the network RTK solution method of the present embodiment includes:

step 1: jointly resolving the baseline and atmosphere (ionosphere and troposphere) delay model parameters involved in the GNSS network according to the atmosphere (ionosphere and troposphere) scale factors,

parameters (such as receiver clock error and carrier phase ambiguity) of various baselines formed between base stations in the network RTK and double-difference atmosphere (ionosphere and troposphere) delay model parameters are jointly solved, namely, the same filter is used for simultaneously estimating all the baseline parameters and the atmosphere parameters. Respectively carrying out random model constraint on ionosphere delay and troposphere delay by using ionosphere scale factors and troposphere scale factors according to a current epoch, wherein the ionosphere scale factor and the troposphere scale factor of a first epoch are set according to historical experience;

step 2: the floating ambiguity closed-loop constraint is that,

adding a double-difference floating ambiguity closed-loop constraint condition, namely setting the closed-difference theoretical value of the double-difference floating ambiguity of each base line in the base station network to be zero, so as to improve the estimation precision of the multi-base-line floating ambiguity solution and provide better initial conditions for integer ambiguity search and fixation;

and step 3: the double-difference ambiguities are searched for over the week,

carrying out integer ambiguity search on the floating ambiguities of all base lines in the base station network, carrying out decorrelation treatment on the floating ambiguities by using the most extensive LAMBDA algorithm, and generating an integer ambiguity candidate subset according to a search range;

and 4, step 4: a whole-cycle ambiguity closed-loop check,

performing ambiguity closed-loop inspection on all the candidates in the whole-cycle ambiguity candidate subset, wherein the ambiguity residual is minimum, and a whole-cycle ambiguity candidate group with which all the satellites conform to the closed-loop inspection is selected as a final ambiguity fixed solution;

and 5: restraining the ambiguity of the whole cycle and solving the atmosphere model

Solving double-difference atmosphere (troposphere and ionosphere) delay of each satellite pair of each baseline by taking a fixed ambiguity solution as a constraint condition, and establishing an atmosphere (ionosphere and troposphere) model by utilizing the solved double-difference atmosphere (ionosphere and troposphere) delay;

the double difference atmosphere (troposphere and ionosphere) delay of each satellite pair of each baseline is calculated and updated, the atmosphere (ionosphere and troposphere) scale factors are used for updating the atmosphere scale factors in the step 1, and the atmosphere (ionosphere and troposphere) delay random model constraint is used in the baseline joint solution of the next epoch;

step 6: virtual reference station VRS station observation data generation

While the preferred embodiments of the present invention have been described in detail, it will be understood by those skilled in the art that the invention is not limited thereto, and that various changes and modifications may be made without departing from the spirit of the invention, and the scope of the appended claims is to be accorded the full scope of the invention.

Claims

1. A network RTK solution method, comprising:

according to the atmospheric scale factor, performing joint calculation on the baseline and atmospheric delay model parameters related in the GNSS reference station network;

setting a double-difference floating-point ambiguity closed-loop constraint condition;

performing integer ambiguity search on the floating ambiguity of each base line in the reference station network together to generate an integer ambiguity candidate subset;

performing ambiguity closed-loop inspection on all the candidates in the integer ambiguity candidate subset, and selecting the integer ambiguity candidate with the minimum ambiguity residual error and all satellites conforming to the closed-loop inspection as a final ambiguity fixed solution;

solving double-difference atmospheric delay of each satellite pair of each base line by taking the ambiguity fixed solution as a constraint condition, and establishing an atmospheric delay model by utilizing the solved double-difference atmospheric delay;

and generating VRS virtual station observation data according to the calculated atmospheric delay model.

2. The method of claim 1, wherein the setting a double-differenced floating ambiguity closed-loop constraint, further comprises:

and setting the closed-range theoretical value of the double-range floating ambiguity of each baseline in the reference station network to be zero.

3. The method of claim 1, wherein when jointly solving the baseline and atmospheric delay model parameters involved in the GNSS network according to the atmospheric scale factor, further comprises:

and constraining a stochastic model of atmospheric delay by the atmospheric scale factor.

4. The method of claim 3, wherein prior to generating VRS virtual station observation data based on the solved atmospheric delay model, further comprising:

and solving the atmospheric scale factor of the current epoch, and constraining the atmospheric delay stochastic model when jointly solving the baseline and atmospheric delay model parameters related to the GNSS reference station network of the next epoch by using the atmospheric scale factor of the current epoch in an epoch real-time feedback manner.

5. The method of claim 1, wherein the jointly solving the baseline parameters and the atmospheric delay model parameters involved in the network of GNSS reference stations according to the atmospheric scale factor further comprises:

and simultaneously estimating all baseline parameters and atmospheric delay model parameters in the GNSS reference station network by using the same filter according to the atmospheric scale factor.

6. The method of claim 1, wherein performing the integer ambiguity search for the floating ambiguities of the baselines in the network of reference stations together to generate the subset of integer ambiguity candidates, further comprises:

and performing integer ambiguity search on the floating ambiguity of each base line in the reference station network, performing decorrelation processing on the floating ambiguity by adopting an LAMBDA algorithm, and generating an integer ambiguity candidate subset according to a search range.

7. The method of claim 4, wherein the initial values of the atmospheric delay model parameters are set based on historical experience.

8. The method of claim 1, wherein the jointly resolving baseline and atmospheric delay model parameters involved in the network of GNSS reference stations further comprises:

and carrying out joint calculation on parameters of various baselines formed among base stations in the reference station network and parameters of a double-difference atmosphere delay model, wherein the parameters of the baselines comprise receiver clock error and carrier ambiguity.

9. The method of any one of claims 1-8, wherein the atmosphere is an ionosphere and a troposphere.

10. The method of claim 3, wherein the constraining the stochastic model of atmospheric delay with the atmospheric scale factor further comprises:

and respectively carrying out random model constraint on the ionosphere delay and the troposphere delay by using the ionosphere scale factor and the troposphere scale factor.