GNSS single-point positioning method and device based on RANSAC, terminal equipment and computer readable storage medium

The RANSAC algorithm is used to iteratively update and judge the observations in GNSS single-point positioning, which solves the problem of error and noise influence in GNSS single-point positioning, and improves the accuracy and robustness of the positioning results.

CN119064971APending Publication Date: 2024-12-03SOUTH SURVEYING & MAPPING INSTR
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411184054.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

In GNSS single-point positioning, the error or noise of the observation value may affect the accuracy of the solution result. The robustness of the traditional method is limited. The chi-square test fails in actual observations, and it is impossible to effectively eliminate observations with large errors.

Method used

Using the RANSAC-based method, the GNSS satellite observation values ​​iteratively updates, randomly selects a subset of observation values ​​to be constructed, calculates the updated parameters to be solved, and judges the inner and outer points based on the observation residuals, and the occupation of the inner points is observed through the observation. The ratio sum-to-average residuals update the parameters, ignoring data points with large errors.

Benefits of technology

It improves the accuracy and robustness of the positioning results, effectively ignores data points containing large errors and noise, and enhances the reliability of the positioning results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119064971A_ABST
    Figure CN119064971A_ABST
Patent Text Reader

Abstract

The invention discloses a GNSS (Global Navigation Satellite System) single-point positioning method and device based on RANSAC (Random Sample Consensus), terminal equipment and a computer readable storage medium. The method comprises the following steps: acquiring a satellite observation value; when the number of satellite observation values is greater than the dimension of the to-be-solved parameter, carrying out iterative updating on the observation values until the number of iterations exceeds a preset number of times, and outputting an optimal receiver coordinate; wherein observation values are randomly selected in each iterative update to form an observation value subset, and an observation equation is constructed for solving to obtain updated to-be-solved parameters; judging whether the number of iterations exceeds a preset number or not, and if yes, outputting the receiver coordinates in the updated to-be-solved parameters; and if not, calculating the observation value residual error, judging whether each observation value is an observation inner point or an observation outer point, and updating the average observation residual error and the parameter solution by comparing the proportion of the observation inner point number with a preset proportion. According to the method and the device, data containing errors and large noise can be ignored, and pollution to a positioning result is avoided.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of satellite navigation and positioning technology, and in particular to a GNSS single-point positioning method based on RANSAC. Background Art

[0002] In GNSS single-point positioning, the error or noise of the observations may affect the accuracy of the solution. To solve this problem, the traditional method usually performs a quality check of the observations after the solution is completed to remove gross error observations. However, the robustness of this method is limited, and some observations with large errors may still be included in the calculation, thus affecting the accuracy of the solution.

[0003] In addition, traditional methods often use chi-square tests to test the conformity of standardized observation residuals. However, this method is based on the assumption that the observations are independent of each other and that the noise conforms to the Gaussian distribution. In actual observations, these assumptions are often not true, resulting in a decrease in the success rate of the chi-square test or a loss of effectiveness. Summary of the invention

[0004] The embodiments of the present invention provide a GNSS single-point positioning method, apparatus, terminal device and computer-readable storage medium based on RANSAC, which can ignore data points containing large errors and noises during the solution process to avoid their contamination of the positioning results.

[0005] An embodiment of the present invention provides a GNSS single point positioning method based on RANSAC, comprising:

[0006] Get several GNSS satellite observations;

[0007] Compare the number of GNSS satellite observations with the dimension of the parameter to be solved;

[0008] When it is determined that the number of GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, the GNSS satellite observation values ​​are iteratively updated by RANSAC until the number of iterations exceeds a preset number, and the optimal receiver coordinates are output;

[0009] In each RANSAC iteration update, GNSS satellite observations with the same dimension as the parameter to be solved and without duplication are randomly selected from the GNSS satellite observations as the current observation subset;

[0010] The observation equation is constructed based on the current observation value subset and the benchmark parameters, and then solved to obtain the updated parameters to be solved; wherein the initial benchmark parameters are 0;

[0011] Determine whether the number of iterations exceeds a preset number, and if so, use the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates;

[0012] If not, the updated average observation residual is calculated based on the observation equation;

[0013] According to the updated parameters to be solved, the residuals of the remaining satellite observation values ​​in the GNSS satellite observation values ​​are calculated, and the observation inner points and the observation outer points are selected according to the residuals of the remaining satellite observation values;

[0014] The proportion of the observed internal point in all the remaining satellite observations is compared with the preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameters to be solved are used as the benchmark parameters for the next iteration; among which, the initial benchmark average observation residual is infinite.

[0015] Furthermore, after comparing the number of GNSS satellite observations with the dimension of the parameter to be solved, it also includes:

[0016] When it is determined that the number of GNSS satellite observation values ​​is equal to the dimension of the parameter to be solved, an observation equation is constructed according to all current GNSS satellite observation values ​​and the initial parameter to be solved, and the equation is solved to calculate the updated parameter to be solved;

[0017] The chi-square test is used to determine whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution. If so, the receiver coordinates in the updated parameter to be solved are taken as the optimal receiver coordinates;

[0018] If not, the standardized residual of each GNSS satellite observation is calculated, and the observation corresponding to the maximum value in the standardized residual is eliminated;

[0019] Re-compare the number of GNSS satellite observations after removing the observations and the dimensions of the parameters to be solved.

[0020] Furthermore, after comparing the number of GNSS satellite observations with the dimension of the parameter to be solved, it also includes:

[0021] When it is determined that the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved, an error signal is output.

[0022] Furthermore, according to the residuals of the remaining satellite observation values, the observation inner points and the observation outer points are selected, including:

[0023] Determine one by one whether the residual of each remaining satellite observation value is less than a preset multiple of the updated average observation residual;

[0024] If so, the corresponding remaining satellite observations are considered as observation inliers;

[0025] If not, the corresponding remaining satellite observations are regarded as observation outliers.

[0026] Furthermore, the preset number of times is determined by the following steps:

[0027] Get the expected RANSAC success rate;

[0028] The preset number of times is calculated using the following formula:

[0029]

[0030] Among them, niter represents the preset number of times, rate represents the expected RANSAC success rate, and npara represents the sum of the dimensions of the parameters in the receiver coordinates to be solved, the dimensions of the satellite system hardware bias parameters, and the dimensions of the ionospheric error parameters. is the floor symbol.

[0031] Furthermore, the observation equation is constructed based on the current observation value subset and the benchmark parameters, and solved to calculate the updated parameters to be solved, including:

[0032] The following first observation equation is constructed based on the current observation value subset and the benchmark parameters:

[0033]

[0034] The first observation equation is linearly expanded to obtain the linearized second observation equation:

[0035]

[0036] V x =X r -X r0 ;

[0037] V y =Y r -Y r0 ;

[0038] V z =Z r -Z r0 ;

[0039] According to the second observation equation corresponding to each observation in the current observation subset, the Jacobian matrix is ​​constructed:

[0040]

[0041] According to the Jacobian matrix, the least square method is used to solve the updated parameters to be solved;

[0042] Among them, X s, Y s and Z s represents the satellite coordinates of the current observation subset, X r , Y r and Z r represents the receiver coordinates in the reference parameters, d sr represents the relative hardware deviation of the satellite system in the reference parameters, V ts represents the satellite clock error, V ion and V trop represent the ionospheric delay and tropospheric delay in the reference parameters, respectively, and X r0 , Y r0 and Z r0 represents the initial receiver coordinates in the benchmark parameters, and H represents the Jacobian matrix.

[0043] Based on the above method embodiment, the present invention provides a corresponding device embodiment, including: a satellite observation value acquisition module, a dimension comparison module, and a solution module;

[0044] Satellite observation value acquisition module, used to obtain several GNSS satellite observation values;

[0045] A dimension comparison module is used to compare the number of GNSS satellite observations with the dimension of the parameter to be solved;

[0046] A solution module, used for performing RANSAC iterative update on the GNSS satellite observation values ​​when it is determined that the number of GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, until the number of iterations exceeds a preset number, and outputting the best receiver coordinates;

[0047] The solution module includes: an observation value subset construction unit, a first observation equation solution unit, a first receiver coordinate output unit, an average observation residual calculation unit, an observation interior point judgment unit, and a data update unit;

[0048] An observation value subset construction unit is used to randomly select GNSS satellite observation values ​​with the same dimension as the parameter to be solved and without duplication from the GNSS satellite observation values ​​in each RANSAC iteration update as the current observation value subset;

[0049] The first observation equation solving unit is used to construct the observation equation according to the current observation value subset and the benchmark parameter, and solve it to calculate the updated parameters to be solved; wherein the benchmark parameter in the first iteration is 0;

[0050] A first receiver coordinate output unit, used to determine whether the number of iterations exceeds a preset number, and if so, to use the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates;

[0051] The average observation residual calculation unit is used to calculate the updated average observation residual according to the observation equation if no;

[0052] An observation interior point judgment unit is used to calculate the residuals of each remaining satellite observation value in the GNSS satellite observation value according to the updated parameters to be solved, and select the observation interior point and the observation exterior point according to the residuals of each remaining satellite observation value;

[0053] The data updating unit is used to compare the proportion of the observed internal point in all the remaining satellite observation values ​​with the preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameter to be solved is used as the benchmark parameter for the next iteration; wherein, the benchmark average observation residual for the first iteration is infinite.

[0054] Further, the solution module further includes: a second observation equation solution unit, a second receiver coordinate output unit, an observation value elimination unit, a dimension re-comparison unit, and an error output unit;

[0055] The second observation equation solving unit is used to construct the observation equation according to all current GNSS satellite observation values ​​and the initial parameters to be solved, and solve the observation equation to calculate the updated parameters to be solved when it is determined that the number of GNSS satellite observation values ​​is equal to the dimension of the parameters to be solved;

[0056] The second receiver coordinate output unit determines whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution through a chi-square test, and if so, takes the receiver coordinates in the updated parameter to be solved as the optimal receiver coordinates;

[0057] an observation value elimination unit, for, if not, calculating the standardized residual of each GNSS satellite observation value in the updated receiver coordinate solution process, and eliminating the observation value corresponding to the maximum value in the standardized residual;

[0058] A dimension re-comparison unit, used to re-comparison the number of GNSS satellite observation values ​​after the observation values ​​are removed and the dimension of the parameter to be solved;

[0059] The error output unit is used to output an error signal when it is determined that the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved.

[0060] Based on the above-mentioned method embodiment, the present invention provides a corresponding terminal device embodiment, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the steps of the GNSS single-point positioning method based on RANSAC as described in the present invention are implemented.

[0061] Based on the above-mentioned method embodiment, the present invention provides a corresponding computer-readable storage medium embodiment, including: a stored computer program, which controls the device where the computer-readable storage medium is located to execute the steps of the GNSS single-point positioning method based on RANSAC as described in the present invention when the computer program is running.

[0062] Compared with the prior art, the beneficial effects of the embodiment of this solution are:

[0063] The present invention obtains a number of GNSS satellite observation values ​​and compares the number of these observation values ​​with the dimension of the parameters to be solved to ensure that the number of GNSS satellite observation values ​​is greater than the dimension of the parameters to be solved. This prerequisite is because in the subsequent solution of the observation equation, at least observation values ​​equal to the dimension of the parameters to be solved are required to form a complete set of equations. If the number of observation values ​​is less than the number of parameters to be solved, sufficient equations cannot be obtained to solve these parameters. Next, RANSAC iterative training is performed on the GNSS satellite observation values ​​until the number of RANSAC iterations exceeds the preset number, and the optimal receiver coordinates are output, thereby obtaining the final positioning coordinates of the receiver. In each RANSAC iterative training, GNSS satellite observations with the same dimension as the parameters to be solved and no duplication are randomly selected from the GNSS satellite observations as the current observation subset; then, the observation equation is constructed based on the observation subset and the reference parameter, and solved to obtain the updated receiver coordinates, where the initial reference parameter is 0; it is determined whether the number of RANSAC iterations exceeds the preset number, if so, the receiver coordinates in the updated parameters to be solved are taken as the optimal receiver coordinates, otherwise, the updated average observation residual is calculated according to the observation equation; according to the updated parameters to be solved, the receiver coordinates are calculated. The residuals of the remaining satellite observations in the GNSS satellite observations are selected, and based on the residuals of the remaining satellite observations, the observation inliers and observation outliers are selected. The observation inliers indicate that the observation is consistent with the estimated parameter solution, and the observation outliers indicate that the observation may contain large errors or noise; the proportion of the observation inliers in all the remaining satellite observations is compared with the preset proportion. If the proportion of the number of observation inliers is greater than the preset proportion, and the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameters to be solved are used as the benchmark parameters for the next iteration. Through the above RANSAC iterative training, the final model outputs the positioning coordinates of the receiver obtained by the observation subset with the most observation inliers. By judging the proportion of observation inliers, data points with large errors and noise are effectively ignored, thereby improving the accuracy of the positioning results.

[0064] In summary, the present invention uses the RANSAC algorithm to determine whether each observation value is an observation inlier or an observation outlier according to the residual of the observation value, and then updates the average observation residual and parameter solution by comparing the proportion of the number of observation inliers with a preset ratio, and selects a subset of observation values ​​with more inliers as the best result, thereby ignoring data points with large errors and noise in the solution process, thereby improving the accuracy and robustness of the positioning results. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1It is a flowchart of a GNSS single point positioning method based on RANSAC provided by an embodiment of the present invention;

[0066] Figure 2 It is a flowchart of an iterative update process of RANSAC provided by an embodiment of the present invention;

[0067] Figure 3 is a structural schematic diagram of a GNSS single-point positioning device based on RANSAC provided by an embodiment of the present invention;

[0068] Figure 4 It is a structural schematic diagram of another GNSS single-point positioning device based on RANSAC provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0070] In the description of the present invention, it should be understood that the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features.

[0071] like Figure 1 As shown, an embodiment of the present invention provides a GNSS single point positioning method based on RANSAC, the method at least comprising the following steps:

[0072] Step S1: Obtain several GNSS satellite observation values;

[0073] For step S1, the receiver receives the GNSS signal transmitted from the air through the antenna and records the pseudo-range observations of each tracking satellite. These observations are important data used by the receiver to calculate its own position and represent the distance information between the receiver and the satellite. The recorded pseudo-range observations will be used for subsequent positioning calculations. They are the key input in the positioning process. By processing and calculating the observations of multiple satellites, the position of the receiver can be derived.

[0074] Step S2: comparing the number of GNSS satellite observations with the dimension of the parameter to be solved;

[0075] For step S2, before using GNSS (Global Navigation Satellite System) for positioning, it is necessary to compare the number of GNSS satellite observations with the dimension of the parameter to be solved, which will result in the following three situations:

[0076] Case 1: If the number of satellite observations is greater than the dimension of the parameters to be solved, it means that we have enough observations to form the observation equations for solution.

[0077] Case 2: If the number of satellite observations is equal to the dimension of the parameters to be solved, it means that we have just enough observations to form the observation equation system for solution.

[0078] Case 3: If the number of satellite observations is less than the dimension of the parameters to be solved, it means that we do not have enough observations to form a complete system of equations to solve. In this case, we cannot solve all the unknown parameters through observations, which means that the position of the receiver cannot be accurately determined.

[0079] It should be noted that the parameters to be solved include the three most basic parameters, the rectangular coordinates x, y, and z of the receiver in the Earth-centered Earth-fixed system, and also include the hardware bias of the satellite system and the ionospheric error. For example, when there is GPS in the observation value, the parameter dimension of the relative hardware bias of the GPS system needs to be added to the parameters to be solved. When there is a Beidou satellite in the observation value, the parameter dimension of the relative hardware bias of the Beidou system needs to be added to the parameters to be solved. Similarly, if there are other satellite systems, it is necessary to add the parameter dimension of the relative hardware bias of other satellite systems. For example, if there is a need to estimate the ionospheric error in the observation value, then the parameter dimension of the corresponding ionospheric error needs to be added to the parameters to be solved.

[0080] Suppose you need to determine the position of a GNSS receiver. The positioning includes three parameters: rectangular coordinates x, y, and z in the Earth-centered Earth-fixed system. For this three-dimensional coordinate position, at least three equations are required to solve these three parameters. If there are only two satellite observations, only two observation equations can be obtained. In this case, the number of equations is less than the number of unknowns, and the accurate position of the receiver cannot be solved. Therefore, before using GNSS for positioning, it is necessary to ensure that the number of satellite observations is at least equal to or more than the dimension of the parameters in the receiver coordinates to be solved in order to obtain accurate positioning results.

[0081] Step S3: when it is determined that the number of GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, the GNSS satellite observation values ​​are iteratively updated by RANSAC until the number of iterations exceeds a preset number, and the optimal receiver coordinates are output;

[0082] In a preferred embodiment, the preset number of times is determined by the following steps:

[0083] Get the expected RANSAC success rate;

[0084] The preset number of times is calculated using the following formula:

[0085]

[0086] Among them, niter represents the preset number of times, rate represents the expected RANSAC success rate, and npara represents the sum of the dimensions of the parameters in the receiver coordinates to be solved, the dimensions of the satellite system hardware bias parameters, and the dimensions of the ionospheric error parameters. is the floor symbol.

[0087] For step S3: when the situation generated after the comparison in step S2 occurs, that is, when it is determined that the number of GNSS satellite observations is greater than the dimension of the parameter to be solved, the observations are updated through RANSAC iteration until the number of iterations reaches a preset number, thereby obtaining the best receiver coordinates. Among them, the preset number of iterations is a value set according to the complexity and accuracy requirements of the problem, and the number set in this embodiment is obtained by a calculation formula.

[0088] Among them, Figure 2 As shown, the iterative update process of RANSAC includes the following steps:

[0089] Step S301: In each RANSAC iterative update, randomly select GNSS satellite observations with the same dimension as the parameter to be solved and without duplication from the GNSS satellite observations as the current observation subset;

[0090] For step S301, in order to reduce the impact of outliers on the estimation results, iteration is performed by randomly selecting subsets. In each RANSAC iteration, a subset of observations with the same dimension as the parameter to be solved is randomly selected from the GNSS satellite observations. This can effectively reduce the impact of outliers and improve the robustness of the estimation. At the same time, it ensures that the selected observations are not repeated in order to avoid using the same data and being unable to solve the receiver coordinates.

[0091] Step S302: constructing an observation equation based on the current observation value subset and the benchmark parameter, and solving it to obtain the updated parameters to be solved; wherein the initial benchmark parameter is 0;

[0092] In a preferred embodiment, an observation equation is constructed based on the current observation value subset and the reference parameter, and is solved to calculate the updated parameters to be solved, including:

[0093] The following first observation equation is constructed based on the current observation value subset and the benchmark parameters:

[0094]

[0095] The first observation equation is linearly expanded to obtain the linearized second observation equation:

[0096]

[0097] V x =X r -X r0 ;

[0098] V y =Y r -Y r0 ;

[0099] V z =Z r -Z r0 ;

[0100] According to the second observation equation corresponding to each observation in the current observation subset, the Jacobian matrix is ​​constructed:

[0101]

[0102] According to the Jacobian matrix, the least square method is used to solve the updated parameters to be solved;

[0103] Among them, X s , Y s and Z s represents the satellite coordinates of the current observation subset, X r , Y r and Z r represents the receiver coordinates in the reference parameters, d sr represents the relative hardware deviation of the satellite system in the reference parameters, V ts represents the satellite clock error, V ion and V trop represent the ionospheric delay and tropospheric delay in the reference parameters, respectively, and X r0 , Y r0 and Z r0 represents the initial receiver coordinates in the benchmark parameters, and H represents the Jacobian matrix.

[0104] For step S302, first, a first observation equation p is constructed according to the current observation value subset and the reference parameter. The observation equation is an equation describing the measurement relationship between the receiver and the satellite. In this embodiment, it represents the observation pseudorange. Then, the first observation equation p is linearly expanded to obtain a linearized second observation equation p ‘The linearized observation equation is the linear equation obtained by performing the first-order Taylor expansion of the nonlinear observation equation at a certain point. The solution of the observation equation is simplified by linearization. Then, according to the linearized observation equation corresponding to each observation value in the current observation value subset, the Jacobian matrix is ​​constructed. The Jacobian matrix is ​​a matrix composed of the partial derivatives of the observation equation with respect to the parameters, which describes the relationship between the observation values ​​and the parameters. Finally, the Jacobian matrix is ​​solved by the least squares method, and the estimated values ​​of the parameters are solved by minimizing the sum of squares of the observation residuals. The estimated values ​​of the parameters are substituted into the observation equation to calculate the updated parameters to be solved.

[0105] Step S303: determining whether the number of iterations exceeds a preset number, and if so, taking the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates;

[0106] For step S303, an increase in the number of iterations will lead to an increase in the amount of calculation. In order to balance the trade-off between calculation efficiency and result accuracy, a reasonable number of iterations has been set through the calculation of the formula. Once this number is exceeded, it can be considered that the result has converged sufficiently, and the receiver coordinates in the updated parameters to be solved are output as the best result.

[0107] Step S304: If not, calculate the updated average observation residual according to the observation equation;

[0108] For step S304, if the number of iterations does not exceed the preset number, after calculating the updated receiver coordinates according to the observation equation, the residual is calculated again using the new coordinates to obtain an updated average observation residual. The average observation residual provides the overall error level between the current predicted position and the actual observation data, which can evaluate the reliability and accuracy of the current predicted receiver coordinates. A smaller average observation residual indicates that the current predicted receiver coordinates are more consistent with the observed data, while a larger average observation residual may indicate that there is a large deviation between the predicted receiver coordinates and the observed data.

[0109] Step S305: Calculate the residuals of the remaining satellite observation values ​​in the GNSS satellite observation values ​​according to the updated parameters to be solved, and select the observation inner points and the observation outer points according to the residuals of the remaining satellite observation values;

[0110] In a preferred embodiment, the observation inliers and outliers are selected according to the residuals of the remaining satellite observation values, including:

[0111] Determine one by one whether the residual of each remaining satellite observation value is less than a preset multiple of the updated average observation residual;

[0112] If so, the corresponding remaining satellite observations are considered as observation inliers;

[0113] If not, the corresponding remaining satellite observations are regarded as observation outliers.

[0114] For step S305, using the receiver coordinates in the updated parameters to be solved, the theoretical value of each satellite observation value can be calculated through the observation equation, and then the actual observation value is subtracted from the theoretical value to obtain the residual of each remaining satellite observation value. At the same time, a preset multiple is set to determine whether the observation value is an observation inlier. In this embodiment, the preset multiple is 1.5 times. Then, for the residual of each satellite observation value, it is determined one by one whether it is less than the preset multiple of the updated average observation residual. If the residual of a certain satellite observation value is less than 1.5 times the updated average observation residual, it means that the observation value is considered to be relatively reliable and contributes to the estimated value in the positioning calculation, and the satellite observation value is regarded as an observation inlier. If the residual of a certain satellite observation value is greater than or equal to 1.5 times the updated average observation residual, it means that the observation value is considered to be possibly affected by abnormalities or unreliable in the positioning calculation of the observation value, which may have a negative impact on the positioning result, and the satellite observation value is regarded as an observation outlier.

[0115] By dividing the observations into observation inliers and observation outliers, it helps to identify abnormal observations that have a negative impact on positioning results and improve the accuracy and robustness of GNSS positioning.

[0116] Step S306: Compare the proportion of the observed inlier point in all remaining satellite observation values ​​with the preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameter to be solved is used as the benchmark parameter for the next iteration; wherein the initial benchmark average observation residual is infinite.

[0117] For step S306, the number of observations determined as observation inliers in all remaining satellite observations is counted, and the proportion of observation inliers in the remaining satellite observations is calculated. Then, the proportion of observation inliers is compared with the preset proportion to determine whether the positioning result has met the expected accuracy requirement. In this embodiment, the preset proportion is 90%. If the proportion of observation inliers is greater than 90%, it means that the receiver coordinates in the parameters to be solved calculated by this RANSAC iteration are calculated using a sufficient number of reliable observations, and if the updated average observation residual is less than the benchmark average observation residual, it means that the current positioning result is more reliable and closer to the true value. Then, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameters to be solved are used as the benchmark parameters for the next iteration.

[0118] By updating the receiver coordinates in the reference parameters, the initial value of the next iteration is made closer to the optimal receiver coordinates, thereby accelerating the convergence process, reducing the number of iterations, and improving the efficiency of the positioning algorithm.

[0119] In addition, it should be noted that the benchmark parameters and benchmark mean observed residuals refer to the parameters to be solved and the mean observed residuals when the previous iteration was updated.

[0120] In a preferred embodiment, after comparing the number of GNSS satellite observation values ​​with the dimension of the parameter to be solved, the method further includes:

[0121] When it is determined that the number of GNSS satellite observation values ​​is equal to the dimension of the parameter to be solved, an observation equation is constructed according to all current GNSS satellite observation values ​​and the initial parameter to be solved, and the equation is solved to calculate the updated parameter to be solved;

[0122] The chi-square test is used to determine whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution. If so, the receiver coordinates in the updated parameter to be solved are taken as the optimal receiver coordinates;

[0123] If not, the standardized residual of each GNSS satellite observation is calculated, and the observation corresponding to the maximum value in the standardized residual is eliminated;

[0124] Re-compare the number of GNSS satellite observations after removing the observations and the dimensions of the parameters to be solved.

[0125] In one embodiment of the present invention, after step S2 compares the number of GNSS satellite observations with the dimension of the parameter to be solved, situation 2 is generated, that is, the number of GNSS satellite observations is equal to the dimension of the parameter to be solved, indicating that there are sufficient observation data to solve the receiver coordinates, and all current satellite observations and the initial parameters to be solved are used to construct the observation equation, and then the observation equation is solved by mathematical methods such as the least squares method to calculate the updated parameters to be solved; the chi-square test is used to evaluate the degree of fit of the updated parameters to be solved, and the chi-square statistic and the critical value of the chi-square distribution are used for comparison. If the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution, it means that the updated parameter to be solved is optimal; if the chi-square statistic of the updated parameter to be solved is not less than the critical value of the chi-square distribution, it means that the updated result is unreliable and there may be abnormal observations. In this case, the standardized residual of each satellite observation is calculated, and the observation corresponding to the maximum value in the standardized residual is eliminated; after eliminating the abnormal observations, the remaining number of satellite observations and the dimension of the parameter to be solved are compared again.

[0126] In a preferred embodiment, after comparing the number of GNSS satellite observation values ​​with the dimension of the parameter to be solved, the method further includes:

[0127] When it is determined that the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved, an error signal is output.

[0128] In one embodiment of the present invention, after step S2 compares the number of GNSS satellite observation values ​​with the dimension of the parameter to be solved, situation three occurs, that is, the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved, indicating that the available observation data is insufficient. Due to the insufficient number of observation values, the system cannot construct a valid observation equation, and therefore cannot perform the subsequent coordinate solution process.

[0129] In this case, the system will recognize that the number of observations is insufficient and output an error signal. This signal indicates that the current observation data cannot support the solution of the receiver coordinates, and the user or the system can take appropriate measures accordingly.

[0130] like Figure 3 As shown, based on the above method embodiment, a corresponding device embodiment is provided;

[0131] An embodiment of the present invention provides a GNSS single-point positioning device based on RANSAC, comprising: a satellite observation value acquisition module, a dimension comparison module, and a solution module;

[0132] Satellite observation value acquisition module, used to obtain several GNSS satellite observation values;

[0133] A dimension comparison module is used to compare the number of GNSS satellite observations with the dimension of the parameter to be solved;

[0134] A solution module, used for performing RANSAC iterative update on the GNSS satellite observation values ​​when it is determined that the number of GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, until the number of iterations exceeds a preset number, and outputting the best receiver coordinates;

[0135] The solution module includes: an observation value subset construction unit, a first observation equation solution unit, a first receiver coordinate output unit, an average observation residual calculation unit, an observation interior point judgment unit, and a data update unit;

[0136] An observation value subset construction unit is used to randomly select GNSS satellite observation values ​​with the same dimension as the parameter to be solved and without duplication from the GNSS satellite observation values ​​in each RANSAC iteration update as the current observation value subset;

[0137] The first observation equation solving unit is used to construct the observation equation according to the current observation value subset and the benchmark parameter, and solve it to calculate the updated parameters to be solved; wherein the benchmark parameter in the first iteration is 0;

[0138] A first receiver coordinate output unit, used to determine whether the number of iterations exceeds a preset number, and if so, to use the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates;

[0139] The average observation residual calculation unit is used to calculate the updated average observation residual according to the observation equation if no;

[0140] An observation interior point judgment unit is used to calculate the residuals of each remaining satellite observation value in the GNSS satellite observation value according to the updated parameters to be solved, and select the observation interior point and the observation exterior point according to the residuals of each remaining satellite observation value;

[0141] The data updating unit is used to compare the proportion of the observed internal point in all the remaining satellite observation values ​​with the preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameter to be solved is used as the benchmark parameter for the next iteration; wherein, the benchmark average observation residual for the first iteration is infinite.

[0142] like Figure 4 As shown, an embodiment of the present invention provides another GNSS single-point positioning device based on RANSAC;

[0143] In a preferred embodiment, the solution module further includes: a second observation equation solution unit, a second receiver coordinate output unit, an observation value elimination unit, a dimension re-comparison unit, and an error output unit;

[0144] The second observation equation solving unit is used to construct the observation equation according to all current GNSS satellite observation values ​​and the initial parameters to be solved, and solve the observation equation to calculate the updated parameters to be solved when it is determined that the number of GNSS satellite observation values ​​is equal to the dimension of the parameters to be solved;

[0145] The second receiver coordinate output unit determines whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution through a chi-square test, and if so, takes the receiver coordinates in the updated parameter to be solved as the optimal receiver coordinates;

[0146] an observation value elimination unit, for, if not, calculating the standardized residual of each GNSS satellite observation value in the updated receiver coordinate solution process, and eliminating the observation value corresponding to the maximum value in the standardized residual;

[0147] A dimension re-comparison unit, used to re-comparison the number of GNSS satellite observation values ​​after the observation values ​​are removed and the dimension of the parameter to be solved;

[0148] The error output unit is used to output an error signal when it is determined that the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved.

[0149] It can be understood that the above-mentioned device item embodiments correspond to the method item embodiments of the present invention, and can implement the RANSAC-based GNSS single-point positioning method provided by any of the above-mentioned method item embodiments of the present invention.

[0150] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art may understand and implement the present invention without creative work.

[0151] Based on the above-mentioned embodiment of the GNSS single-point positioning method based on RANSAC, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the GNSS single-point positioning method based on RANSAC according to any embodiment of the present invention is implemented.

[0152] Exemplarily, in this embodiment, the computer program may be divided into one or more modules, which are stored in the memory and executed by the processor to implement the present invention. The one or more module elements may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the terminal device.

[0153] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0154] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0155] Based on the above method embodiment, another embodiment is provided: another embodiment of the present invention provides a computer-readable storage medium, including a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the RANSAC-based GNSS single-point positioning method described in any one of the above method embodiments of the present invention.

[0156] Wherein, the module / unit integrated in the GNSS single-point positioning device / terminal device based on RANSAC, if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on such an understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Wherein, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0157] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A GNSS single point positioning method based on RANSAC, characterized in that: include: Get several GNSS satellite observations; Compare the number of GNSS satellite observations with the dimension of the parameter to be solved; When it is determined that the number of the GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, performing RANSAC iterative update on the GNSS satellite observation values ​​until the number of iterations exceeds a preset number, and outputting the optimal receiver coordinates; In each RANSAC iteration update, GNSS satellite observations with the same dimension as the parameter to be solved and without duplication are randomly selected from the GNSS satellite observations as the current observation subset; The observation equation is constructed based on the current observation value subset and the benchmark parameters, and then solved to obtain the updated parameters to be solved; wherein the initial benchmark parameters are 0; Determine whether the number of iterations exceeds a preset number, and if so, use the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates; If not, the updated average observation residual is calculated based on the observation equation; According to the updated parameters to be solved, the residuals of the remaining satellite observation values ​​in the GNSS satellite observation values ​​are calculated, and the observation inner points and the observation outer points are selected according to the residuals of the remaining satellite observation values; The proportion of the observed internal point in all the remaining satellite observations is compared with the preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameters to be solved are used as the benchmark parameters for the next iteration; among which, the initial benchmark average observation residual is infinite.

2. The GNSS single point positioning method based on RANSAC according to claim 1, characterized in that: After comparing the number of GNSS satellite observations with the dimensions of the parameters to be solved, it also includes: When it is determined that the number of the GNSS satellite observation values ​​is equal to the dimension of the parameter to be solved, constructing an observation equation according to all current GNSS satellite observation values ​​and the initial parameter to be solved, and solving the equation to calculate and obtain an updated parameter to be solved; Determine whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution through a chi-square test, and if so, take the receiver coordinates in the updated parameter to be solved as the optimal receiver coordinates; If not, the standardized residual of each GNSS satellite observation is calculated, and the observation corresponding to the maximum value in the standardized residual is eliminated; Re-compare the number of GNSS satellite observations after removing the observations and the dimensions of the parameters to be solved.

3. The GNSS single point positioning method based on RANSAC according to claim 2, characterized in that: After comparing the number of GNSS satellite observations with the dimensions of the parameters to be solved, it also includes: When it is determined that the number of the GNSS satellite observation values ​​is less than the dimension of the parameter to be solved, an error signal is output.

4. The GNSS single point positioning method based on RANSAC according to claim 3, characterized in that: According to the residuals of the remaining satellite observation values, the observation inner points and observation outer points are selected, including: Determine one by one whether the residual of each remaining satellite observation value is less than a preset multiple of the updated average observation residual; If so, the corresponding remaining satellite observations are considered as observation inliers; If not, the corresponding remaining satellite observations are regarded as observation outliers.

5. The GNSS single point positioning method based on RANSAC according to claim 3, characterized in that: The preset number of times is determined by the following steps: Get the expected RANSAC success rate; The preset number of times is calculated by the following formula: Among them, niter represents the preset number of times, rate represents the expected RANSAC success rate, and npara represents the sum of the dimensions of the parameters in the receiver coordinates to be solved, the dimensions of the satellite system hardware bias parameters, and the dimensions of the ionospheric error parameters. is the floor symbol.

6. The GNSS single point positioning method based on RANSAC according to claim 3, characterized in that: The observation equation is constructed based on the current observation value subset and the benchmark parameters, and then solved to obtain the updated parameters to be solved, including: The following first observation equation is constructed based on the current observation value subset and the benchmark parameters: The first observation equation is linearly expanded to obtain the linearized second observation equation: V x =X r -X r0 ; V y =And r -AND r0 ; In z =Z r -Z r0 ; According to the second observation equation corresponding to each observation in the current observation subset, the Jacobian matrix is ​​constructed: According to the Jacobian matrix, the least square method is used to solve the updated parameters to be solved; Among them, X s , Y s and Z s represents the satellite coordinates of the current observation subset, X r , Y r and Z r represents the receiver coordinates in the reference parameters, d sr represents the relative hardware deviation of the satellite system in the reference parameters, V ts represents the satellite clock error, V ion and V trop represent the ionospheric delay and tropospheric delay in the reference parameters, respectively, and X r0 , Y r0 and Z r0 represents the initial receiver coordinates in the benchmark parameters, and H represents the Jacobian matrix.

7. A GNSS single-point positioning device based on RANSAC, characterized in that: include: Satellite observation value acquisition module, dimension comparison module, and solution module; The satellite observation value acquisition module is used to acquire a number of GNSS satellite observation values; The dimension comparison module is used to compare the number of GNSS satellite observation values ​​with the dimension of the parameter to be solved; The solving module is used to perform RANSAC iterative update on the GNSS satellite observation values ​​when it is determined that the number of the GNSS satellite observation values ​​is greater than the dimension of the parameter to be solved, until the number of iterations exceeds a preset number, and output the optimal receiver coordinates; The solution module includes: an observation value subset construction unit, a first observation equation solution unit, a first receiver coordinate output unit, an average observation residual calculation unit, an observation interior point judgment unit, and a data update unit; The observation value subset construction unit is used to randomly select GNSS satellite observation values ​​that have the same dimension as the parameter to be solved and are non-repetitive from the GNSS satellite observation values ​​as the current observation value subset in each RANSAC iterative update; The first observation equation solving unit is used to construct the observation equation according to the current observation value subset and the reference parameter, and solve it to calculate the updated parameters to be solved; wherein the reference parameter in the first iteration is 0; The first receiver coordinate output unit is used to determine whether the number of iterations exceeds a preset number, and if so, to use the updated receiver coordinates in the parameters to be solved as the optimal receiver coordinates; The average observation residual calculation unit is used to calculate the updated average observation residual according to the observation equation if no; The observation inlier judgment unit is used to calculate the residuals of the remaining satellite observation values ​​in the GNSS satellite observation values ​​according to the updated parameters to be solved, and select the observation inlier points and the observation outlier points according to the residuals of the remaining satellite observation values; The data updating unit is used to compare the proportion of the observed internal point in all the remaining satellite observation values ​​with a preset proportion; if it is greater than the preset proportion, then when the updated average observation residual is less than the benchmark average observation residual, the updated average observation residual is used as the benchmark average observation residual for the next iteration, and the updated parameter to be solved is used as the benchmark parameter for the next iteration; wherein the benchmark average observation residual for the first iteration is infinite.

8. The GNSS single-point positioning device based on RANSAC according to claim 7, characterized in that: The solution module further includes: a second observation equation solution unit, a second receiver coordinate output unit, an observation value elimination unit, a dimension re-comparison unit, and an error output unit; The second observation equation solving unit is used to construct an observation equation according to all current GNSS satellite observation values ​​and the initial parameters to be solved, and solve the observation equation to calculate the updated parameters to be solved when it is determined that the number of the GNSS satellite observation values ​​is equal to the dimension of the parameters to be solved; The second receiver coordinate output unit is used to determine whether the chi-square statistic of the updated parameter to be solved is less than the critical value of the chi-square distribution through a chi-square test, and if so, the receiver coordinates in the updated parameter to be solved are used as the optimal receiver coordinates; The observation value elimination unit is used to calculate the standardized residual of each GNSS satellite observation value if no, and eliminate the observation value corresponding to the maximum value in the standardized residual; The dimension re-comparison unit is used to re-comparison the number of GNSS satellite observation values ​​after the observation values ​​are removed and the dimension of the parameter to be solved; The error output unit is used to output an error signal when it is determined that the number of GNSS satellite observation values ​​is less than the dimension of the parameter to be solved.

9. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the RANSAC-based GNSS single-point positioning method as described in any one of claims 1 to 6 is implemented.

10. A computer-readable storage medium, characterized in that: include: A stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the RANSAC-based GNSS single-point positioning method according to any one of claims 1 to 6.

Citation Information

Cited By

  • Method for estimating observed quantity noise of GNSS (Global Navigation Satellite System) receiver

    CN119355765A

  • GNSS coordinate displacement detection method and system combined with long and short time window segmentation

    CN119687767A

  • Receiver high-precision positioning method and system based on RANSAC algorithm

    CN120722406A

  • High-precision positioning method and system for receiver based on ransac algorithm

    CN120722406B