A positioning solution method and system based on Kalman filtering

Abstract

The application provides a positioning solution method and system based on Kalman filtering, and relates to the technical field of global satellite navigation and positioning. The method comprises the following steps: receiving satellite original data and reference station differential data, and constructing a double-difference observation error equation; performing positioning solution on the double-difference observation error equation based on Kalman filtering; and determining the position information of the current state based on the positioning solution result. In the positioning solution process, the state transition matrix and the observation matrix of Kalman filtering are rearranged and combined, and are decomposed into blocks comprising multiple special matrices, each block is operated separately, the calculation amount is reduced, the storage space occupied in the positioning solution process is reduced, and the positioning solution efficiency is improved.

Description

Technical Field

[0001] This application belongs to the field of global satellite navigation and positioning technology, and in particular relates to a positioning solution method and system based on Kalman filtering. Background Technology

[0002] The statements in this section are merely background information related to this application and do not necessarily constitute prior art.

[0003] Real-time dynamic differential (RTK) is a relative positioning method where the rover (user) processes observations from two stations in real time relative to a reference station. The implementation of an RTK positioning terminal relies heavily on chips and embedded systems, requiring significant computational resources. This includes processing high-sampling-frequency data for fixed-solution output, and involves Kalman filter estimation of multi-dimensional state parameters. This estimation process involves numerous high-dimensional floating-point matrix multiplication operations. Furthermore, the low cost and performance of chips, coupled with the limited CPU and resources of embedded systems, restrict RTK computation efficiency. Therefore, as the amount of satellite observations increases, the positioning efficiency based on Kalman filtering remains relatively low. Summary of the Invention

[0004] To address the aforementioned issues, the first aspect of this application provides a positioning solution method based on Kalman filtering. By rearranging and combining the Kalman filter state transition matrix and observation matrix, the method decomposes the matrix into blocks comprising multiple special matrices. Each block is processed separately to reduce computational load and storage space occupied during positioning, thereby improving positioning efficiency.

[0005] To achieve the above objectives, this application mainly includes the following aspects:

[0006] In a first aspect, embodiments of this application provide a localization solution method based on Kalman filtering, including:

[0007] Receive raw satellite data and differential data from reference stations, and construct a double-difference observation error equation;

[0008] The double-difference observation error equation is solved using Kalman filtering. During the solution process, the state transition matrix is ​​divided into blocks, each of which is processed independently to obtain the state vector and state variance-covariance matrix for one step prediction. The observation matrix is ​​rearranged and decomposed into blocks containing multiple special matrices, each of which is processed independently. The Kalman gain matrix is ​​determined based on the block-wise observation matrix and state variance-covariance matrix. The state vector and state variance-covariance matrix of the current epoch are further updated based on the Kalman gain matrix.

[0009] Based on the positioning calculation results, the current position information is determined.

[0010] In one possible implementation, the observation matrix is ​​rearranged and combined based on preset rules, which include: shifting column vectors in the same row of the observation matrix that all contain zeros to the left or right, or shifting row vectors in the same column of the observation matrix that all contain zeros to the top or bottom.

[0011] In one possible implementation, the special matrix includes a zero matrix, an identity matrix, and a diagonal matrix.

[0012] In one possible implementation, the state transition matrix is ​​rearranged and combined to decompose it into blocks comprising multiple special matrices.

[0013] In one possible implementation, determining the Kalman gain matrix based on the segmented observation matrix and the state variance-covariance matrix includes: decomposing the state variance-covariance matrix into blocks corresponding to the observation matrix, and determining the Kalman gain matrix based on the segmented observation matrix and the segmented state variance-covariance matrix.

[0014] In one possible implementation, determining the current state's location information based on the positioning solution results includes: estimating the user's state parameters using Kalman filtering to obtain a floating-point solution, then performing an integer ambiguity search to obtain a fixed solution, thereby determining the current state's location information. The state parameters include position, velocity, and ambiguity.

[0015] Secondly, embodiments of this application provide a positioning solution system based on Kalman filtering, comprising:

[0016] The receiving module is used to receive raw satellite data and differential data from the reference station, and to construct the double-difference observation error equation;

[0017] The positioning solution module is used to perform positioning solutions on the double-difference observation error equation based on Kalman filtering. During the positioning solution process, the state transition matrix is ​​divided into blocks, and each block is calculated separately to obtain the state vector and state variance-covariance matrix for one step prediction. The observation matrix is ​​rearranged and combined into blocks containing multiple special matrices, and each block is calculated separately. The Kalman gain matrix is ​​determined based on the block-wise observation matrix and state variance-covariance matrix. The state vector and state variance-covariance matrix of the current epoch are further updated based on the Kalman gain matrix.

[0018] The location information determination module is used to determine the current location information based on the positioning calculation results.

[0019] In one possible implementation, the positioning calculation module is used to rearrange and combine the observation matrix based on preset rules, the preset rules including: shifting the column vectors of the observation matrix whose rows all contain zero to the left or right, or shifting the row vectors of the observation matrix whose columns all contain zero to the right or left.

[0020] Thirdly, embodiments of this application provide a computer device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the computer device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of the localization solution method based on Kalman filtering as described in the first aspect and any possible implementation of the first aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the localization solution method based on Kalman filtering as described in the first aspect and any possible implementation thereof.

[0022] The beneficial effects of this application are:

[0023] 1. This application aims to improve the efficiency of localization by rearranging and combining the Kalman filter state transition matrix and observation matrix into blocks containing multiple special matrices, and performing separate operations on each block, thereby reducing the amount of computation and the storage space occupied during the localization process.

[0024] 2. The Kalman filter state transition matrix and observation matrix are rearranged and combined according to preset rules to perform matrix block decomposition, decomposing multiple special matrix blocks including zero matrix, identity matrix or diagonal matrix. The appearance of zero matrix and identity matrix can eliminate redundant calculations, and the appearance of diagonal matrix can further reduce the number of matrix multiplication loops. Thus, in matrix block operation, not only is the dimension of the matrix reduced, but the number of matrix multiplications is also greatly reduced, thereby greatly improving the RTK solution efficiency.

[0025] 3. When the terminal performs RTK positioning, it uses general Kalman filtering to estimate state parameters. The amount of floating-point matrix operation increases with the increase of satellite observations. Low-cost chips or embedded systems will limit the RTK solution efficiency, making single-epoch RTK solution take more processing time and unable to reach the expected processing frequency. Using the method provided in this application for embedded RTK positioning, the solution efficiency can be improved by about 30% compared with the former, which makes up for the impact of insufficient chip performance and greatly improves the RTK solution efficiency. Enterprises can not only develop high-efficiency and high-reliability RTK positioning terminals without increasing costs, but also reduce hardware costs to a certain extent. Attached Figure Description

[0026] The accompanying drawings, which form part of this application, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0027] Figure 1 This is one of the flowcharts illustrating the localization solution method based on Kalman filtering provided in this embodiment of the invention;

[0028] Figure 2 This is the second flowchart of the localization solution method based on Kalman filtering provided in the embodiments of this application;

[0029] Figure 3 This is a schematic diagram of a positioning solution system based on Kalman filtering provided in an embodiment of this application;

[0030] Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0031] The present application will be further described below with reference to the accompanying drawings and embodiments.

[0032] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0033] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0034] Real-time dynamic differential (RTK) technology is a relative positioning method in which the rover (user) processes the observations of two stations in real time relative to a reference station. Its main idea is to send differential data such as carrier phase, pseudorange, Doppler, and signal-to-noise ratio collected by the reference station to the user, while the user receives the raw satellite data. By performing a second differential between the observations, most of the satellite-end error, receiver-end error, and atmospheric delay error are eliminated. A Kalman filter is then used to estimate the user's position, velocity, ambiguity, and other state parameters to obtain a floating-point solution. Then, an integer ambiguity search is performed to obtain a fixed solution, and finally, the precise position information of the terminal is obtained.

[0035] Kalman filtering involves high-dimensional floating-point matrix multiplication, and the dimension increases with the number of available satellites. Assuming atmospheric delay residual estimation is not considered, under observation conditions with n satellites, when calculating using dual-frequency carrier and pseudorange measurements, the number of state parameters is (including 2n ambiguity parameters, position parameters in three directions, velocity parameters in three directions, and acceleration parameters in three directions); the state transition matrix size is (nx × nx); the design matrix (observation matrix) size is (nv × nx), where nv = 4n; and the state variance-covariance matrix size is (nx × nx). For low-cost, low-performance chips, the matrix computation load increases exponentially with increasing dimension, significantly impacting RTK processing efficiency. If development requires high-frequency RTK fixed-solution output greater than 1Hz, it may even lead to situations where untimely processing of current epoch data affects the RTK solution for the next epoch, resulting in module output loss, stuttering, and system crashes.

[0036] To address the issue of reduced efficiency in Kalman filtering due to increased observations and to improve RTK solution efficiency, this application proposes a positioning solution method based on Kalman filtering. This method divides the state transition matrix and observation matrix into blocks, performing separate calculations on each block to reduce computational load and storage space usage during positioning solution, thereby improving positioning solution efficiency.

[0037] Example 1

[0038] Please see Figure 1 and Figure 2 The localization solution method based on Kalman filtering includes the following steps:

[0039] S101: Receives raw satellite data and differential data from the reference station, and constructs a double-difference observation error equation.

[0040] The reference station differential data includes carrier phase, pseudorange, Doppler, and signal-to-noise ratio.

[0041] S102: The double-difference observation error equation is solved based on Kalman filtering. During the localization process, the state transition matrix is ​​divided into blocks, and each block is calculated separately to obtain the state vector and state variance-covariance matrix for one step prediction. The observation matrix is ​​rearranged and combined into blocks containing multiple special matrices, and each block is calculated separately. The Kalman gain matrix is ​​determined based on the block-wise observation matrix and state variance-covariance matrix. The state vector and state variance-covariance matrix of the current epoch are further updated based on the Kalman gain matrix.

[0042] In specific implementation, before introducing the localization solution method based on Kalman filtering described in this application, the general process of Kalman filtering calculation is given first. Kalman filtering includes the following calculation process:

[0043] The one-step prediction of the state vector and state variance-covariance matrix is ​​expressed as:

[0044]

[0045]

[0046] In the formula, the subscripts k-1 and k represent the identifiers of the previous epoch and the current epoch in the prediction process, respectively; X is the predicted state vector for the current epoch; k-1 is the state vector of the previous epoch; F is the state transition matrix; u represents the external influence, which is the zero matrix here; P is the state variance-covariance matrix predicted for the current epoch; k-1 is the state variance-covariance matrix of the previous epoch; Q is the process noise variance matrix.

[0047] The update of the state vector and the state variance-covariance matrix is ​​expressed as follows:

[0048]

[0049]

[0050]

[0051] In the formula, X k P k The updated state vector and state variance-covariance matrix of the current epoch, which take into account the predicted values ​​and observations, can be used for Kalman filter state prediction in the next epoch; K is the Kalman gain; V is the double-difference observation residual vector; H is the single-difference observation matrix; R is the a priori variance-oblique variance matrix; and I is the identity matrix.

[0052] This application optimizes the one-step state prediction and state update during the localization solution of the double-difference observation error equation based on Kalman filtering. In one embodiment of this application, as an optional implementation, the observation matrix is ​​rearranged and combined according to preset rules. These preset rules include: shifting column vectors in the same row of the observation matrix that all contain zeros left and right, or shifting row vectors in the same column of the observation matrix that all contain zeros up and down. Optionally, the special matrices include a zero matrix, an identity matrix, and a diagonal matrix; the state transition matrix is ​​rearranged and combined, decomposing it into blocks including multiple special matrices; determining the Kalman gain matrix based on the block-wise observation matrix and the state variance-covariance matrix includes: decomposing the state variance-covariance matrix into blocks corresponding to the observation matrix, and determining the Kalman gain matrix based on the block-wise observation matrix and the block-wise state variance-covariance matrix. In specific applications, the following steps are included:

[0053] A. One-step state prediction optimization

[0054] In the one-step prediction of the state vector and state variance-covariance matrix, considering that the state transition matrix F is a sparse matrix, it is partitioned into blocks to decompose the zero matrix and the identity matrix. The state transition matrix F is partitioned as follows:

[0055]

[0056] This can be further expressed as:

[0057]

[0058] In the formula, F 12 and F 21 All are zero matrices, F 22 Since F is the identity matrix, the state transition matrix F can be simplified to the following form:

[0059]

[0060] First, substituting the above equation into the one-step prediction formula for the state vector, we get:

[0061]

[0062] Further simplification yields the optimized one-step prediction formula for the state vector, as follows:

[0063]

[0064] In the formula, U is the state estimation parameter vector for position, velocity, and acceleration; N is the single-difference floating-point ambiguity state estimation parameter vector, including the frequency point 1 ambiguity state estimation parameter vector N1 and the frequency point 2 ambiguity state estimation parameter vector N2; nx is the number of estimated state parameters. Therefore, the optimized state vector prediction formula only needs to calculate F. 11 .U, which is a multiplication operation of a 9-dimensional square matrix and a 9-dimensional vector, significantly reduces the number of matrix multiplications compared to the previous multiplication of an nx-dimensional square matrix and an nx-dimensional vector, thus significantly improving the RTK solution efficiency. For matrix F... 11 Further analysis reveals that the matrix is ​​an upper triangular matrix. While matrix multiplication can be simplified by increasing the number of multiplications, its size is only 9×9, so the efficiency improvement after optimization is minimal and may not be necessary. This will not be elaborated further here.

[0065] Then, before substituting the F matrix into the one-step prediction formula for the state variance-covariance matrix, first calculate the state variance-covariance matrix P... k-1 Decomposed into 2×2 matrix blocks, due to the limitation of the operating system's bit width, P accumulates over time. k-1 The matrix will lose some floating-point precision, resulting in matrix P k-1 No longer symmetric, making matrix P k-1 Since it has lost its symmetry property and is not a sparse matrix, it cannot be reorganized for optimization. Therefore, we will not perform an expansion analysis here, but can obtain the following results by dividing it into blocks:

[0066]

[0067] F and P k-1 Substituting the state variance-covariance matrix into the one-step prediction formula yields:

[0068]

[0069] The simplified formula for the one-step prediction of the state variance-covariance matrix is ​​as follows:

[0070]

[0071] As can be seen from the formula, the dimensionality of the optimized state variance-covariance matrix one-step prediction formula is greatly reduced, leaving only some multiplication operations between low-dimensional matrices, and avoiding many redundant multiplication operations, thus significantly reducing the amount of computation.

[0072] B. Status Update Optimization

[0073] In updating the state vector and the state variance-covariance matrix, considering that the observation matrix H is a sparse matrix, its column vectors are adjusted left and right, and the matrix is ​​partitioned into blocks to decompose it into a zero matrix and a diagonal matrix. The observation matrix H is represented as follows:

[0074]

[0075] In the formula, (e x ,e y ,e z () represents the line-of-sight vector between the receiver and the satellite. This indicates the carrier phase observations of the reference satellite r in system S; This indicates the carrier phase observation flag for the first satellite in system S; nvc is the total number of carrier phase observations used. This indicates the pseudorange observation marker for the reference star r in system S; This indicates the pseudorange observation flag for the first satellite in system S; nvp is the total number of pseudorange observations used. The carrier wavelength is the first satellite in system S. It should be noted that the carrier phase and pseudorange observations include single-frequency or multi-frequency data, which are not discussed here.

[0076] The observation matrix H is further adjusted as follows:

[0077]

[0078] Furthermore, the observation matrix H can be decomposed into 2×3 matrix blocks, represented as follows:

[0079]

[0080] In the formula, H 12 The matrix is ​​a diagonal matrix, which reduces the number of loops and multiplications during matrix multiplication; n1 is the total number of parameters including position, reference star floating-point ambiguity, and state parameters; nx satisfies nx = n1 + nvc + 6; nv satisfies nv = nvc + nvp; after block division, three zero matrices H are generated. 13 H 22 H 23 This greatly reduces the number of matrix multiplications.

[0081] Additionally, the state variance-covariance matrix needs to be... Decomposed into 3×3 matrix blocks, due to the matrix There are no special characteristics, so we will not elaborate on the analysis here. We can divide the data into blocks and obtain the following:

[0082]

[0083] First, H and Substituting into the Kalman gain equation, let:

[0084]

[0085] Therefore, the optimized Kalman gain equation is:

[0086]

[0087]

[0088] Then, the state variance-covariance matrix update formula is optimized by letting:

[0089]

[0090] Therefore, the optimized state variance-covariance matrix update formula is:

[0091]

[0092] The optimized Kalman gain equation and state variance-covariance matrix update formula are slightly more complex in expression than the unoptimized ones, but after optimization, except for H... 12 Outside the matrix, only some low-dimensional matrix multiplication operations remain, and after removing a large number of zero matrices, many redundant multiplication operations are avoided, significantly reducing the computational load; H 12 Although the dimension is still relatively large, it is a diagonal matrix. When performing matrix multiplication, it is only necessary to traverse the diagonal elements, which greatly reduces the amount of computation required for matrix multiplication.

[0093] The optimized state vector update formula is expressed as follows: While the expression is similar to the general formula, the optimized Kalman gain formula can significantly reduce the computational cost of matrix multiplication and improve the efficiency of RTK solution.

[0094] S103: Determine the current location information based on the positioning solution results.

[0095] As an optional implementation, the location information of the current state is determined based on the positioning solution results, including: estimating the user's state parameters based on Kalman filtering to obtain a floating-point solution, and then performing integer ambiguity search to obtain a fixed solution, thereby determining the location information of the current state, wherein the state parameters include position, velocity, and ambiguity.

[0096] The localization solution method based on Kalman filtering provided in this application aims to improve the efficiency of localization by rearranging and combining the Kalman filter state transition matrix and observation matrix into blocks containing multiple special matrices, with each block processed independently. This reduces computational load and storage space usage during localization. Furthermore, the Kalman filter state transition matrix and observation matrix are rearranged and combined according to preset rules to perform matrix block decomposition, resulting in multiple special matrix blocks including zero matrices, identity matrices, or diagonal matrices. The presence of zero matrices and identity matrices eliminates redundant computations, while the presence of diagonal matrices further reduces the number of matrix multiplication loops. Therefore, during matrix block operations, not only is the matrix dimension reduced, but the number of matrix multiplications is also significantly decreased, which greatly improves the efficiency of RTK solution.

[0097] Example 2

[0098] Please see Figure 3 , Figure 3 This is the localization solution system based on Kalman filtering provided in the embodiments of this application, such as... Figure 3 As shown in the figure, this application embodiment provides a positioning solution system 300 based on Kalman filtering, including:

[0099] Receiver module 310 is used to receive raw satellite data and differential data from the reference station to construct the double-difference observation error equation;

[0100] The positioning solution module 320 is used to perform positioning solution on the double-difference observation error equation based on Kalman filtering. During the positioning solution process, the state transition matrix is ​​divided into blocks, and each block is calculated separately to obtain the state vector and state variance-covariance matrix for one step prediction. The observation matrix is ​​rearranged and combined to decompose it into blocks including multiple special matrices, and each block is calculated separately. The Kalman gain matrix is ​​determined based on the block-wise observation matrix and state variance-covariance matrix. The state vector and state variance-covariance matrix of the current epoch are further updated based on the Kalman gain matrix.

[0101] The location information determination module 330 is used to determine the location information of the current state based on the positioning calculation results.

[0102] In this embodiment of the application, as an optional embodiment, the positioning calculation module is used to rearrange and combine the observation matrix based on preset rules. The preset rules include: shifting the column vectors of the observation matrix whose rows all contain zero to the left or right, or shifting the row vectors of the observation matrix whose columns all contain zero to the right or left.

[0103] The positioning and solving system based on Kalman filtering provided in this application includes a receiving module, a positioning and solving module, and a location information determination module. By rearranging and combining the Kalman filter state transition matrix and the observation matrix, it decomposes them into blocks containing multiple special matrices. Each block is processed independently to reduce computational load and storage space occupied during positioning and solving, thereby improving positioning and solving efficiency. Furthermore, the Kalman filter state transition matrix and the observation matrix are rearranged and combined according to preset rules to perform matrix block decomposition, resulting in multiple special matrix blocks including zero matrices, identity matrices, or diagonal matrices. The presence of zero matrices and identity matrices eliminates redundant calculations, and the presence of diagonal matrices further reduces the number of matrix multiplication loops. Therefore, during matrix block operations, not only is the matrix dimension reduced, but the number of matrix multiplications is also greatly reduced, which will significantly improve RTK solving efficiency.

[0104] Furthermore, when the terminal performs RTK positioning, the floating-point matrix operation required for estimating state parameters using general Kalman filtering increases with the amount of satellite observations. Low-cost chips or embedded systems can limit the RTK solution efficiency, causing single-epoch RTK solutions to take more processing time and fail to achieve the expected processing frequency. Using the method provided in this application for embedded RTK positioning, the solution efficiency can be improved by about 30% compared to the former, compensating for the impact of insufficient chip performance and greatly improving RTK solution efficiency. Enterprises can not only develop high-efficiency and high-reliability RTK positioning terminals without increasing costs, but also reduce hardware costs to a certain extent.

[0105] Example 3

[0106] Please see Figure 4 , Figure 4 This is a schematic diagram of a computer device according to an embodiment of this application. Figure 4 As shown, the computer device 400 includes a processor 410, a memory 420, and a bus 430.

[0107] The memory 420 stores machine-readable instructions executable by the processor 410. When the computer device 400 is running, the processor 410 communicates with the memory 420 via the bus 430. When the machine-readable instructions are executed by the processor 410, they can perform the operations described above. Figure 1 and Figure 2 The steps of the localization solution method based on Kalman filtering in the method embodiment shown are described in detail in the method embodiment, and will not be repeated here.

[0108] Example 4

[0109] Based on the same concept, embodiments of this application also provide a computer-readable storage medium storing a computer program, which, when run by a processor, executes the steps of the localization solution method based on Kalman filtering described in the above method embodiments.

[0110] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0111] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A Kalman filter based positioning solution method, characterized in that, The method comprises the following steps: Receiving satellite raw data and reference station differential data, and constructing a double-difference observation error equation; Based on Kalman filtering, the double-difference observation error equation is solved, and in the process of positioning solution, the state transition matrix is matrix blocked, each block is operated separately, and a one-step predicted state vector and a state variance-covariance matrix are obtained; The observation matrix is rearranged and combined, decomposed into blocks including multiple special matrices, each block is operated separately, the Kalman gain matrix is determined according to the blocked observation matrix and the state variance-covariance matrix, and the state vector and the state variance-covariance matrix of the current epoch are further updated according to the Kalman gain matrix; The observation matrix is rearranged and combined based on a preset rule, and the preset rule comprises: moving the column vectors of the same row of the observation matrix containing zeros left and right, or moving the row vectors of the same column of the observation matrix containing zeros up and down; The state transition matrix is rearranged and combined, and decomposed into blocks including multiple special matrices; Based on the positioning solution result, the position information of the current state is determined.

2. The Kalman filter based positioning solution method of claim 1, wherein, The special matrix comprises a zero matrix, a unit matrix and a diagonal matrix.

3. The Kalman filter based positioning solution method of claim 1, wherein, The Kalman gain matrix is determined according to the blocked observation matrix and the state variance-covariance matrix, which comprises: the state variance-covariance matrix is decomposed into blocks corresponding to the observation matrix, and the Kalman gain matrix is determined according to the blocked observation matrix and the blocked state variance-covariance matrix.

4. The Kalman filter based positioning solution method of claim 1, wherein, Based on the positioning solution result, the position information of the current state is determined, which comprises: based on Kalman filtering, the state parameters of the user are estimated to obtain a float solution, then a whole-week ambiguity search is performed to obtain a fixed solution, and the position information of the current state is determined, wherein the state parameters comprise position, velocity and ambiguity.

5. A Kalman filter based positioning solution system employing the method of claim 1, characterized by The method comprises the following steps: A receiving module is configured to receive satellite raw data and reference station differential data, and construct a double-difference observation error equation; A positioning solution module is configured to solve the double-difference observation error equation based on Kalman filtering, and in the process of positioning solution, the state transition matrix is matrix blocked, each block is operated separately, and a one-step predicted state vector and a state variance-covariance matrix are obtained; The observation matrix is rearranged and combined, decomposed into blocks including multiple special matrices, each block is operated separately, the Kalman gain matrix is determined according to the blocked observation matrix and the state variance-covariance matrix, and the state vector and the state variance-covariance matrix of the current epoch are further updated according to the Kalman gain matrix; A position information determination module is configured to determine the position information of the current state based on the positioning solution result.

6. The Kalman filter based positioning solution system of claim 5, wherein, The positioning solution module is configured to rearrange and combine the observation matrix based on a preset rule, and the preset rule comprises: moving the column vectors of the same row of the observation matrix containing zeros left and right, or moving the row vectors of the same column of the observation matrix containing zeros up and down.

7. A computer device, comprising: The method comprises the following steps: A processor, a memory, and a bus, the memory storing machine readable instructions executable by the processor, when the computer device is running, the processor and the memory communicating through the bus, the machine readable instructions being executed by the processor to perform the steps of the Kalman filter based positioning method according to any one of claims 1 to 4.

8. A computer-readable storage medium, characterized in that, A computer readable storage medium storing a computer program, the computer program being executed by the processor to perform the steps of the Kalman filter based positioning method according to any one of claims 1 to 4.

Images