Adaptive GNSS position parameter state updating method based on elastic thermodynamic grid search

Abstract

This invention discloses an adaptive GNSS position parameter state update method based on elastic thermal grid search; belonging to the field of GNSS positioning and navigation technology; the method includes: starting a differential relative positioning program, storing fixed solution coordinates and their covariance information in real time; performing a discreteness test on the accumulated historical fixed epochs, and when the fixed solution distribution meets the stability condition, taking the current accumulated number of fixed solutions as the total number of valid samples; using the initial fixed solution of the station as the initial heatmap center, adaptively determining the heatmap search range and step size, constructing a multi-layer thermal grid for layer-by-layer refinement search until the layered search converges, and outputting the coordinate parameters and their covariance corresponding to the center of the grid with the highest heat in the final layer; re-initializing the position parameters with the coordinate parameters and their covariance as initial variables, and continuing to execute differential relative positioning until the observation data solution is completed. This invention suppresses multipath errors and improves the solution accuracy of fixed position parameters.

Description

Technical Field

[0001] This invention relates to the field of GNSS positioning and navigation technology; and more particularly to an adaptive GNSS position parameter state update method based on thermal grid search. Background Technology

[0002] Landslides, mudslides, and other geological disasters cause serious harm to people's lives and livelihoods. High-precision monitoring and early warning can effectively prevent these geological disasters from causing losses to people's lives and property. GNSS (Global Navigation Satellite System) differential relative positioning technology has advantages such as low cost, high accuracy, and all-weather operation, and is being increasingly deployed in areas prone to geological disasters for deformation monitoring and early warning.

[0003] While GNSS differential relative positioning technology can eliminate the influence of errors such as tropospheric delay and ionospheric delay by taking advantage of short baselines, the accuracy of fixed position parameter solutions is still affected by interference from multipath errors. In addition, the geological change characteristics of different monitoring stations are different, and it is not accurate to update the state parameters by using pseudo-dynamic or static position parameter state transition models.

[0004] Therefore, considering the impact of factors such as multipath error and complex deformation patterns in GNSS deformation monitoring, methods to improve the solution accuracy of fixed position parameters have become an industry demand. Summary of the Invention

[0005] Based on the above analysis, this invention aims to disclose an adaptive GNSS position parameter state update method based on thermal grid search; to suppress multipath error and improve the solution accuracy of fixed position parameter solutions.

[0006] This invention discloses an adaptive GNSS position parameter state update method based on elastic thermal grid search, comprising: S1. Start the differential relative positioning program, obtain the floating-point parameter solution and fix the ambiguity, and store the fixed solution coordinates and their covariance information in real time. S2. Perform a dispersion test on the accumulated historical fixed epochs. When the distribution of fixed solutions meets the stability condition, take the current accumulated number of fixed solutions as the total number of valid samples and trigger thermal grid modeling. S3. Using the initial fixed solution of the station as the initial heat map center, the heat map search range and step size are adaptively determined based on the historical maximum three-dimensional uncertainty and the total number of effective samples. A multi-layer thermal grid is constructed to refine the search layer by layer until the layered search converges. The coordinate parameters and covariance of the grid center with the highest heat intensity in the final layer are output. S4. Reinitialize the position parameters using the coordinate parameters and their covariance as initial variables, return to S1 to continue differential relative positioning until the observation data solution is completed.

[0007] Furthermore, the convergence of the hierarchical search is determined based on the differences in search results between adjacent layers, the degree of improvement in heat distribution, and / or the current search accuracy.

[0008] Furthermore, the difference in search results between adjacent layers is determined according to the first formula: First formula: ; when When the condition for convergence is met, the first condition for convergence is determined, where, The preset spatial distance threshold; , The current layer The next layer The center coordinates of the thermal grid; The degree of improvement in heat distribution is determined according to the second formula: Second formula: ; when When the second convergence condition is met, it is determined that the convergence condition is satisfied, where, The preset proportional change threshold; , Do not use the current layer The next layer The grid heat of the hottest grid; The number of valid samples; The current search precision is determined according to the third formula: ; If the third formula holds, the third convergence condition is satisfied, where... For the current layer Step size, This represents the upper limit of the minimum step size.

[0009] Furthermore, the preset spatial distance threshold Based on the current layer search step size Determined proportionally; ; This is the proportionality coefficient. 3~0.6; The preset ratio change threshold The determination is adaptively based on the total number of valid samples and the distribution characteristics of historical fixed solutions.

[0010] Furthermore, in S2, the dispersion of the spatial distribution of the fixed solution coordinates is quantitatively evaluated to perform a dispersion test; The stability condition is as follows: the dispersion of the fixed solution coordinates is calculated in real time, and the fixed solution distribution is considered to be stable when the dispersion is less than a threshold. The dispersion calculation includes: for n fixed solution coordinates, calculating the average coordinates in the three directions of east, north, and sky respectively; calculating the deviation of each fixed solution coordinate in the corresponding direction based on the average coordinates in each direction; and calculating the three-dimensional dispersion index based on the deviation.

[0011] Furthermore, S3 includes: S3-1. Using the initial fixed solution of the station as the center of the initial heat map, the search range and step size of the heat map are adaptively determined based on the historical uncertainty. S3-2. Using the current heatmap center as the search center, perform grid division within the search range, determine the grid heat based on the fixed number of solutions falling into each grid, and identify the grid with the highest heat. S3-3. Calculate the center coordinates of the next thermal grid based on the fixed solution coordinates within the grid with the highest heat intensity. S3-4. Determine whether the layered search of the multi-layer thermal grid has converged. If it has, proceed to step S3-6; otherwise, proceed to step S3-5. S3-5. Using the grid with the highest heat value as the next search area, and the center coordinates of the next layer as the new search center, reduce the search step size and return to step S3-2. S3-6. Take the average of the fixed solution coordinates in the grid with the highest heat in the final layer to obtain the coordinate parameters, take the average of the corresponding covariances to obtain the covariance parameters, and output the coordinate parameters and their covariances.

[0012] Furthermore, the heatmap search range for: ; in, The maximum uncertainty for a fixed solution coordinate is... k Confidence coefficient; Step length for: ; in, The minimum step size threshold, This represents the resolution coefficient.

[0013] Furthermore, the search step size is reduced according to a preset ratio.

[0014] Furthermore, the differential relative positioning program is initiated, and multi-mode GNSS parameters are estimated based on extended Kalman filtering to obtain floating-point position parameters and integer ambiguity floating-point solutions, while the corresponding state covariance matrix is ​​estimated. The LAMBDA model is used to fix the integer ambiguity floating-point solution to obtain the fixed solution coordinates and their covariance information; The fixed solution coordinates and their covariance information are stored in the form of a hash table.

[0015] Compared with traditional methods, the present invention has the following technical advantages: The adaptive GNSS position parameter state update method based on elastic thermal grid search of the present invention addresses the problem of difficulty in accurately estimating station coordinates in fields with slowly changing environments, such as deformation monitoring containing multipath errors. It introduces a thermal grid search model and combines it with the Ratio fixed rate to search for the position parameter with the highest accuracy. During the state transition process, it performs accurate position parameter updates, thereby improving the accuracy of station coordinate calculation. Attached Figure Description

[0016] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts. Figure 1 This is a flowchart of the adaptive GNSS position parameter state update method based on elastic thermal grid search in an embodiment of the present invention; Figure 2 This is the first thermal grid diagram when Layer=2 in one embodiment of the present invention; Figure 3 This is the second thermal grid diagram when one layer = 2 in this embodiment of the invention; Figure 4 The adaptive GNSS position parameter state update method for force grid search in this embodiment of the invention uses the E-direction coordinate sequence diagram before and after; Figure 5 The adaptive GNSS position parameter state update method for thermal grid search in this embodiment of the invention uses the preceding and following N-direction coordinate sequence diagrams; Figure 6 The adaptive GNSS position parameter state update method for thermal grid search in this embodiment of the invention uses the U-direction coordinate sequence diagram before and after. Detailed Implementation

[0017] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and, together with the embodiments of the present invention, serve to illustrate the principles of the present invention.

[0018] Example 1 One embodiment of the present invention discloses an adaptive GNSS position parameter state update method based on elastic thermal grid search, such as... Figure 1 As shown, it includes: S1. Start the differential relative positioning program, obtain the floating-point parameter solution and fix the ambiguity, and store the fixed solution coordinates and their covariance information in real time. S2. Perform a dispersion test on the accumulated historical fixed epochs. When the distribution of fixed solutions meets the stability condition, take the current accumulated number of fixed solutions as the total number of valid samples and trigger thermal grid modeling. S3. Using the initial fixed solution of the station as the initial heat map center, the heat map search range and step size are adaptively determined based on the historical maximum three-dimensional uncertainty and the total number of effective samples. A multi-layer thermal grid is constructed to refine the search layer by layer until the layered search converges. The coordinate parameters and covariance of the grid center with the highest heat intensity in the final layer are output. S4. Reinitialize the position parameters using the coordinate parameters and their covariance as initial variables, return to S1 to continue differential relative positioning until the observation data solution is completed.

[0019] Specifically, in S1, Start the differential relative positioning program, perform multi-mode GNSS parameter estimation based on extended Kalman filtering, obtain floating-point position parameters and integer ambiguity floating-point solutions, and estimate the corresponding state covariance matrix. The least squares ambiguity decorrelation adjustment (LAMBDA) model is used to fix the integer ambiguity floating-point solution, and the fixed solution coordinates and their covariance information are obtained. The fixed solution coordinates and their covariance information are stored in the form of a hash table.

[0020] In this embodiment, both the extended Kalman filter and the LAMBDA model can be derived from technical solutions disclosed in existing literature.

[0021] The fixed solution coordinates and their covariance information are stored in the form of a hash table; the hash table uses epoch time as the key value, supports fast retrieval and reading of fixed solution information, and provides a data foundation for subsequent thermal grid modeling.

[0022] Specifically, in S2, The dispersion test is achieved by quantitatively evaluating the degree of dispersion of the spatial distribution of fixed solution coordinates; The stability condition is as follows: the dispersion of the fixed solution coordinates is calculated in real time, and when the dispersion is less than a threshold, the fixed solution distribution is considered stable. The dispersion calculation includes: for n fixed solution coordinates, calculating the average coordinates in the east, north, and sky directions respectively; calculating the deviation of each fixed solution coordinate in the corresponding direction based on the average coordinates in each direction; and calculating the three-dimensional dispersion index based on the deviation. The three-dimensional dispersion index is the root mean square or arithmetic mean of the standard deviations in each direction. The threshold is determined based on the product of the historical maximum three-dimensional uncertainty and a proportionality coefficient, wherein the proportionality coefficient is preferably between 0.3 and 0.5. When the fixed solution distribution does not meet the stability condition, continue to accumulate the fixed solution and recalculate the dispersion until the stability condition is met or the maximum cumulative epoch is reached. The maximum cumulative epoch number is dynamically adjusted according to the observation sampling rate and application scenario, and is preferably set to 5-10 minutes of equivalent epoch number. When the maximum cumulative epoch number is reached but the stability condition is still not met, the current cumulative number of fixed solutions is used as an effective sample to force the thermal grid modeling, and the confidence coefficient is increased in subsequent solutions to expand the search range and compensate for the model uncertainty caused by the unstable distribution of fixed solutions.

[0023] Specifically, S3 includes: S3-1. Using the initial fixed solution of the station as the center of the initial heat map, the search range and step size of the heat map are adaptively determined based on the historical uncertainty. Using the initial fixed solution of the station as the center of the initial heat map, the search range and step size of the heat map are adaptively determined based on the maximum uncertainty σ of the fixed solution coordinates. When the search is first triggered, the historical statistical calculation is based on all fixed solutions accumulated in S1-S2. The calculation method can be the maximum value or root mean square of the standard deviation components of the floating-point position parameter covariance matrix output by the extended Kalman filter in the three directions of east, north, and sky, or the historical maximum value of the three-dimensional deviation between the fixed solution coordinates and the reference coordinates. When subsequent iterations trigger the search, the coordinate parameter covariance is updated based on the output of the previous S3-S4 solution, or a sliding window is used to statistically update the most recent fixed solution. The adaptively determined heatmap search range for: ; in, The maximum uncertainty of the fixed solution coordinates is given by k, which is the confidence coefficient. It is preferably 2-4. When the observation quality decreases or the fixation rate decreases, a larger value is taken to expand the search range. Step length for: ; in, The minimum step size threshold, The resolution coefficient is preferably set to 3-6.

[0024] In the preferred scheme, the confidence coefficient can be dynamically adjusted. When the observation environment deteriorates, leading to a decrease in observation quality or a reduction in the fixation rate, the confidence coefficient can be appropriately increased. The value of is chosen to expand the search range of the thermal grid, thereby encompassing more outlier fixed solutions that may be affected by multipath errors, and improving the robustness of the thermal grid to complex observation environments; conversely, when observation conditions are good and the fixation rate is stable, a smaller value can be chosen. This value is used to improve search efficiency.

[0025] S3-2. Using the current heatmap center as the search center, perform grid division within the search range, determine the grid heat based on the fixed number of solutions falling into each grid, and identify the grid with the highest heat. The grid is divided into three-dimensional grids with equal spacing along the east, north, and sky directions within the search range, with the current heat map center as the origin. Each grid is identified by the coordinates of its center. The grid heat is the number of fixed solutions falling into the grid, or the ratio of the number of fixed solutions falling into the grid to the total number of valid samples. The process of identifying the grid with the highest heat index includes: traversing all grids, comparing the heat index of each grid, and taking the grid with the highest heat index as the grid with the highest heat index; when there are multiple grids with the same heat index, taking the grid whose geometric center is closest to the center of the current heat map, or taking the merged area of ​​grids with the same heat index. The adaptively determined mesh partitioning step size is the interval in S3-1; The number of grids within the search range is: ; Where range is the search range determined by S3-1, and interval is the step size determined by S3-1; When there are many fixed solutions or a large search range, hash mapping or spatial indexing can be used to accelerate grid heat statistics.

[0026] S3-3. Calculate the center coordinates of the next thermal grid based on the fixed solution coordinates within the grid with the highest heat intensity. Calculate the center coordinates of the next thermal grid based on the fixed solution coordinates within the grid with the highest thermal intensity. The calculation involves taking the arithmetic mean of the coordinates of all fixed solutions within the grid with the highest heat intensity, and using the three-dimensional average coordinates as the center of the next layer. The center coordinates of the next layer are: ; in, This represents the fixed number of solutions within the grid with the highest current heat level. For the first in this grid The three-dimensional coordinates of a fixed solution; When the fixed solutions are unevenly distributed within the grid with the highest heat, a weighted average is used. The weights are determined based on the covariance or Ratio test value of each fixed solution. The smaller the covariance or the larger the Ratio value, the greater the weight.

[0027] S3-4. Determine whether the layered search of the multi-layer thermal grid has converged. If it has, proceed to step S3-6; otherwise, proceed to step S3-5. Among them, the convergence of hierarchical search can be determined based on the differences in search results between adjacent layers, the degree of improvement in heat distribution, and / or the current search accuracy; The difference in search results between adjacent layers is determined according to the first formula: First formula: ; when When the condition for convergence is met, the first condition for convergence is determined, where, The preset spatial distance threshold; , The current layer The next layer The center coordinates of the thermal grid; The degree of improvement in heat distribution is determined according to the second formula: Second formula: ; when When the second convergence condition is met, it is determined that the convergence condition is satisfied, where, The preset proportional change threshold; , Do not use the current layer The next layer The grid heat of the hottest grid; The number of valid samples; The current search precision is determined according to the third formula: ; If the third formula holds, the third convergence condition is satisfied, where... For the current layer Step size, This is the upper limit of the minimum step size, which can be adaptively set according to the observation scenario.

[0028] The preset spatial distance threshold Based on the current layer search step size Determined proportionally; ; This is the proportionality coefficient. 3~0.6; The preset ratio change threshold The value is adaptively determined based on the total number of valid samples and the distribution characteristics of historical fixed solutions. When the distribution of fixed solutions is highly discrete, a larger value is taken to improve the convergence tolerance, and when the distribution of fixed solutions is concentrated, a smaller value is taken to improve the convergence accuracy. The preferred value is 0.05-0.15. In practical use, any one, any two, or all of the three conditions can be selected to determine the convergence of the hierarchical search and stop building the heatmap. The convergence of the hierarchical search is determined as soon as any condition is met. This is the simplest to implement, with low computational cost and fast convergence speed. However, the convergence accuracy is relatively limited. The search may terminate prematurely due to the local satisfaction of a single condition, resulting in the final output coordinate parameters not being close enough to the optimal solution. The refinement stops when there are slight fluctuations in the fixed solution distribution, and the spatial resolution of the heatmap is insufficient. The convergence of hierarchical search is determined when all conditions are met simultaneously. This ensures that the heatmap meets the expected standards in three dimensions: center stability, heat concentration, and search granularity. The output results are more accurate and reliable. However, the convergence conditions are too stringent. Especially when the fixed solution distribution is discrete or the observation quality fluctuates, the difficulty of satisfying multiple conditions at the same time increases significantly. It is easy to fail to converge, which may cause the search level to exceed the limit or fall into an infinite loop, affecting the solution efficiency and real-time performance. However, the fulfillment of any two conditions can serve as a compromise strategy; Furthermore, to address the issue of convergence failing based on conditions, when the preset maximum search level is reached, the hierarchical search is forcibly terminated, and steps S3-6 are initiated.

[0029] S3-5. Using the grid with the highest heat value as the next search area, and the center coordinates of the next layer as the new search center, reduce the search step size and return to step S3-2. Among them, the search step size is reduced according to a preset ratio; The next layer search region does not exceed the boundary range of the hottest grid to ensure the continuity and convergence of the search.

[0030] S3-6. Take the average of the fixed solution coordinates in the grid with the highest heat in the final layer to obtain the coordinate parameters, take the average of the corresponding covariances to obtain the covariance parameters, and output the coordinate parameters and their covariances.

[0031] The coordinate parameters are obtained by averaging the fixed solution coordinates within the grid with the highest heat in the final layer, and the covariances are obtained by averaging the corresponding covariances. The coordinate parameters and their covariances are then output. The final layer is either the current layer that meets the termination condition S3-4, or the last layer that reaches the preset maximum search level. The coordinate parameters are the arithmetic mean of the three-dimensional coordinates of all fixed solutions within the grid with the highest heat in the final layer: ; The covariance parameter is the arithmetic mean of the corresponding fixed solution covariance matrix: ; in, This represents the fixed number of solutions within the grid with the highest heat in the final layer. For the first in this grid The three-dimensional coordinates of a fixed solution This is the corresponding covariance matrix; When the number of fixed solutions in the final layer is insufficient, backtrack to the layer that most recently met the quantity requirement, or expand the search range of that layer to include fixed solutions from adjacent grids to ensure statistical reliability.

[0032] Specifically, in S4, state updates and iterative solutions are performed; The coordinate parameters and their covariance are used as initial variables to reinitialize the position parameters. Then, return to step S1 to continue performing differential relative positioning until the observation data solution is completed. The re-initialization includes: using the coordinate parameters As the initial value of the position parameter in the Kalman filter state vector, the covariance parameter is used. The initial values ​​of the state covariance of the corresponding position parameters are used to reset the state estimate and covariance matrix of the filter. The return step S1 includes: retaining the current observation data and satellite ephemeris information, and continuing the differential relative positioning solution for the next epoch based on the re-initialized state parameters; The calculation end judgment includes: judging whether the current observation time has reached the last time of the data sequence, or judging whether the calculation time has reached the preset monitoring period. If the conditions are met, the calculation is terminated and all results are output; otherwise, iterative execution continues.

[0033] In summary, the adaptive GNSS position parameter state update method based on elastic thermal grid search in this embodiment addresses the problem of inaccurate estimation of station coordinates in slowly changing environments, such as deformation monitoring with multipath errors. It introduces a thermal grid search model and combines it with the Ratio fixed rate to search for the position parameter with the highest accuracy. During the state transition process, it performs precise position parameter updates, thereby improving the accuracy of station coordinate calculation.

[0034] Example 2 This embodiment uses data from two GNSS multi-system continuously operating observation stations approximately 18m apart as an example to monitor the dynamic deformation of a building, and specifically illustrates the application process of this invention.

[0035] We selected 24-hour observation data for 334 days in 2021 as the target data, and set the dispersion test threshold and the uncertainty of fixed solution coordinates. Heatmap search hierarchy (Layer), fixed total number of samples Grid search range With step size After configuring the thermal grid parameters, single-epoch differential relative positioning calculations are performed, and the LAMBDA model is used to fix the ambiguity.

[0036] The historical fixed epoch is judged based on the dispersion test model. If the index is met, the construction of the elastic thermal grid is triggered, and the total number of valid samples is counted. .

[0037] Using the initial fixed solution of the station as the center of the initial heatmap, the search range of the heatmap is determined based on the historical maximum three-dimensional uncertainty model. The value is , where the confidence coefficient Maximum uncertainty m, resolution coefficient Step length It is 0.02m.

[0038] Set criteria for changes in the center coordinates of adjacent layers. Establish a Layer criterion for elastic thermal grid search, and obtain results such as... Figure 2 , Figure 3 The diagram shows a two-layer thermal grid. Figure 2 This is the first layer of thermal grid diagram. Figure 3 The second layer of thermal grid diagram shows that, after layer-by-layer refinement, the distribution of fixed solutions is more concentrated.

[0039] When the hierarchical search stops, output the coordinate parameters corresponding to the center of the highest heatmap grid in the final heatmap. and its covariance When the position parameter status is updated, and The position parameters are reinitialized as initial variables, and the above process is repeated until the solution is completed.

[0040] Get as Figure 4 , Figure 5 , Figure 6 The figure shows the coordinate deviation sequences of the east, north, and sky directions using and without a thermal grid model. Red represents the coordinate sequence obtained using the thermal grid model, and black represents the coordinate sequence obtained without it. As can be seen from the figure, after applying the method of this invention, the dispersion of the coordinate sequence is significantly reduced, the influence of multipath error is effectively suppressed, and the accuracy of the position parameter calculation is significantly improved.

[0041] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An adaptive GNSS position parameter state update method based on elastic thermal grid search, characterized in that, include: S1. Start the differential relative positioning program, obtain the floating-point parameter solution and fix the ambiguity, and store the fixed solution coordinates and their covariance information in real time. S2. Perform a dispersion test on the accumulated historical fixed epochs. When the distribution of fixed solutions meets the stability condition, take the current accumulated number of fixed solutions as the total number of valid samples and trigger thermal grid modeling. S3. Using the initial fixed solution of the station as the initial heat map center, the heat map search range and step size are adaptively determined based on the historical maximum three-dimensional uncertainty and the total number of effective samples. A multi-layer thermal grid is constructed to refine the search layer by layer until the layered search converges. The coordinate parameters and covariance of the grid center with the highest heat intensity in the final layer are output. S4. Reinitialize the position parameters using the coordinate parameters and their covariance as initial variables, return to S1 to continue differential relative positioning until the observation data solution is completed; S3 includes: S3-1. Using the initial fixed solution of the station as the center of the initial heat map, the search range and step size of the heat map are adaptively determined based on the historical uncertainty. The heatmap search range for: ； in, The maximum uncertainty for a fixed solution coordinate is... k Confidence coefficient; Step length for: ； in, The minimum step size threshold, This refers to the resolution coefficient; S3-2. Using the current heatmap center as the search center, perform grid division within the search range, determine the grid heat based on the fixed number of solutions falling into each grid, and identify the grid with the highest heat. The grid heat is the ratio of the number of fixed solutions falling into the grid to the total number of valid samples; S3-3. Calculate the center coordinates of the next thermal grid based on the fixed solution coordinates within the grid with the highest heat intensity. S3-4. Determine whether the layered search of the multi-layer thermal grid has converged. If it has, proceed to step S3-6; otherwise, proceed to step S3-5. Among them, the convergence of hierarchical search is determined based on the differences in search results between adjacent layers, the degree of improvement in heat distribution, and / or the current search accuracy; The difference in search results between adjacent layers is determined according to the first formula: First formula: ; when When the condition for convergence is met, the first condition for convergence is determined, where, The preset spatial distance threshold; , The current layer The next layer The center coordinates of the thermal grid; The degree of improvement in heat distribution is determined according to the second formula: Second formula: ; when When the second convergence condition is met, it is determined that the convergence condition is satisfied, where, The preset proportional change threshold; , Do not use the current layer The next layer The grid heat of the hottest grid; The number of valid samples; The current search precision is determined according to the third formula: ； If the third formula holds, the third convergence condition is satisfied, where... For the current layer Step size, This represents the upper limit of the minimum step size; The preset spatial distance threshold Based on the current layer search step size Determined proportionally; ； This is the proportionality coefficient. 3~0.6; The preset ratio change threshold The determination is adaptively based on the total number of valid samples and the distribution characteristics of historical fixed solutions; The value is set to 0.05-0.

15. When the fixed solution distribution is highly discrete, a larger value is used to improve the convergence tolerance, and when the fixed solution distribution is concentrated, a smaller value is used to improve the convergence accuracy. S3-5. Using the grid with the highest heat value as the next search area, and the center coordinates of the next layer as the new search center, reduce the search step size and return to step S3-2. S3-6. Take the average of the fixed solution coordinates in the grid with the highest heat in the final layer to obtain the coordinate parameters, take the average of the corresponding covariances to obtain the covariance parameters, and output the coordinate parameters and their covariances.

2. The adaptive GNSS position parameter state update method based on elastic thermal grid search according to claim 1, characterized in that, In S2, the dispersion of the spatial distribution of the fixed solution coordinates is quantitatively evaluated to perform a dispersion test. The stability condition is as follows: the dispersion of the fixed solution coordinates is calculated in real time, and the fixed solution distribution is considered to be stable when the dispersion is less than a threshold. The dispersion calculation includes: for n fixed solution coordinates, calculating the average coordinates in the three directions of east, north, and sky respectively; calculating the deviation of each fixed solution coordinate in the corresponding direction based on the average coordinates in each direction; and calculating the three-dimensional dispersion index based on the deviation.

3. The adaptive GNSS position parameter state update method based on elastic thermal grid search according to claim 1, characterized in that, Reduce the search step size according to the preset ratio.

4. The adaptive GNSS position parameter state update method based on elastic thermal grid search according to claim 1, characterized in that, Start the differential relative positioning program, perform multi-mode GNSS parameter estimation based on extended Kalman filtering, obtain floating-point position parameters and integer ambiguity floating-point solutions, and estimate the corresponding state covariance matrix. The LAMBDA model is used to fix the integer ambiguity floating-point solution to obtain the fixed solution coordinates and their covariance information; The fixed solution coordinates and their covariance information are stored in the form of a hash table.

Images