VRS grid division method

Abstract

The invention provides a VRS grid division method. The method comprises the following steps: calculating a regional atmospheric delay gradient based on historical observation data of a CORS base station; calculating a settable grid VRS maximum service spacing and a settable grid VRS minimum service spacing in the area; dividing N types of grid spacing intervals; performing equal-interval grid division according to longitude and latitude grids; grid atmosphere delay and effective service spacing are calculated; calculating a maximum height difference and a height difference influence factor in the grid; calculating time-space characteristics of users in the grid; calculating a grid spacing scaling factor; the grid spacing is calculated; n types of grid intervals are matched and clustered; and grid coding and edge matching processing. According to the VR S grid division method, the influence of multiple factors such as the maximum height difference in the grid, the user space-time characteristic and the effective service spacing of the grid is considered, and the influence degree of different factors on the grid division effect is considered for weighted calculation. Compared with the prior art, the method has the advantages that the division is more accurate and fine, the server VRS calculation pressure and the moving station positioning effect are better balanced, and the positioning effect is greatly improved.

Description

Technical Field

[0001] The present application relates to the technical field of high-precision location services, and in particular to a VRS grid division method. Background Art

[0002] Since the positioning error has spatial correlation. For example, the closer the mobile station is to the VRS station, the higher the positioning accuracy of the mobile station. Therefore, compared with the ordinary VRS technology, the grid-based VRS technology reduces the computing pressure on the server by sacrificing a certain positioning accuracy. The denser the grid, the higher the positioning accuracy, but the computing pressure on the server is greater. On the contrary, the sparser the grid, the lower the positioning accuracy, and the computing pressure on the server is less. Therefore, the VRS grid division method directly affects the user's positioning accuracy and the computing pressure on the server, and is a key issue that needs to be solved by the grid-based VRS technology.

[0003] At present, mainstream CORS data processing software sets the virtual grid spacing through empirical values, which cannot achieve the optimal balance between user positioning accuracy and server-side data solution pressure. In addition, due to differences in geographical location, observation period, terrain environment, and atmospheric delay, there are differences in services across grids.

[0004] Li Ying et al. analyzed the error sources of grid VRS, verified the interpolation effect of the interpolation model with additional elevation correction, and studied the appropriate density of grid division using the ground-based augmentation system in Henan Province as an example. Chen Chong et al. proposed a method to generate irregular grids to fit the terrain. According to the digital elevation model information in the region, an irregular grid that matches the terrain in the region is interpolated. Huang Dingfa et al. proposed a gridded VRS method for wide-area concurrent user services, and studied the effective service spacing of gridded VRS under different terrain conditions. Shi Mingxing proposed a gridded VRS division method based on the analysis of the spatiotemporal characteristics of atmospheric delay, and designed a method for determining the virtual grid spacing and dynamically evaluating service accuracy.

[0005] However, the VRS grid division methods of the above scholars only take into account the terrain, or take into account a few factors such as terrain and atmospheric delay, and directly delineate the grid in the region, and still complete the grid pre-division based on the empirical model. There are the following three defects:

[0006] First, the influence of different factors on the grid division effect is not considered, and a comprehensive weighted evaluation is not conducted. Only a single influencing factor is analyzed, and the grid division effect has a large room for improvement;

[0007] Second, the actual spatiotemporal distribution characteristics of user mobile stations and the real-time positioning effect are not considered, and the grid division strategy is dynamically iterated and adjusted. The empirical model completes the grid pre-division, which still causes a waste of computing resources on the server side.

[0008] Third, it fails to take into account the complex terrain, where irregular grid division will lead to grid coding confusion, which is not conducive to grid management and rapid matching of user-side mobile station grids. Summary of the invention

[0009] In order to solve at least one of the technical problems in the background technology, the embodiment of the present application provides a VRS grid division method, comprising the following steps:

[0010] Step S1, calculating the regional atmospheric delay gradient based on the historical observation data of the CORS reference station, and converting it into error gradients in the N and E directions;

[0011] Step S2, setting a positioning error threshold, and calculating the maximum grid VRS service spacing and the minimum grid VRS service spacing that can be set in the area based on the atmospheric delay gradient;

[0012] Step S3, according to the grid VRS maximum service spacing and the grid VRS minimum service spacing calculated in step S2, the grid spacing is set to N types of intervals in an arithmetic progression;

[0013] Step S4, dividing the area into equal intervals according to longitude and latitude grids using the grid spacing corresponding to rounding N / 2;

[0014] Step S5: Calculate the atmospheric delay and effective service spacing D in each latitude and longitude grid according to the regional atmospheric delay gradient calculated in step S1 and the positioning error threshold set in step S2. i ;

[0015] Step S6: Sampling is performed at a spacing of 1 km × 1 km, sampling is performed by fitting the digital elevation model data in the area, and the maximum height difference ΔH in the grid is calculated. i , and set the height difference threshold H, and calculate the height difference influence factor as follows:

[0016]

[0017] Step S7: Calculate the spatiotemporal characteristic index of users in the grid using historical user GGA data. The calculation formula is as follows:

[0018]

[0019] In the formula, UI i is the user spatiotemporal characteristic index of grid i, P USI is the user spatial activity weight, which can be adjusted according to experience or actual test data, n is the number of all users in grid i during the statistical time, N is the total number of grids, and m is the number of users in all grids in the region during the statistical time; P UAIis the user's spatiotemporal availability weight, which can be adjusted based on experience or actual test data. t is the sampling time, which is 1-s and is expressed in days. t is the number of fixed users in grid i on the sampling day, n t is the number of all users in grid i on the sampling day;

[0020] Step S8: Maximum height difference EH in the grid i , User spatiotemporal characteristic index UI i , grid effective service spacing D i The three factors are weighted to calculate the grid spacing scaling factor, where the scaling factors in the N direction and the E direction can be calculated separately. The calculation formula is as follows:

[0021]

[0022] In the formula, SF i is the scaling factor, D is the latitude and longitude grid spacing, P D is the effective service distance weight, P UI is the user's spatiotemporal feature weight, is the maximum height difference weight within the grid;

[0023] Step S9: Divide the latitude and longitude grid spacing D by the scaling factor SF i , get the final spacing of each grid;

[0024] Step S10: combining the final spacing of each grid in step S9 and the N-type grid spacing division result in step S3, classifying all grids, using DBSCAN algorithm, selecting the empirical clustering threshold, and dividing the entire area into N grid clustering spaces;

[0025] Step S11, encode the longitude and latitude of the clustered grids, and check whether there are service gaps in the grid VRS at the edges of different cluster spaces. If there are service gaps, add grid VRS at the edges to ensure full service coverage.

[0026] In one possible implementation manner, in the VRS grid division method provided in the embodiment of the present application, in the step S2, the positioning error threshold is not more than 2 cm in both the N and E directions.

[0027] In one possible implementation manner, the VRS grid division method provided in the embodiment of the present application, in the step S2, the grid VRS maximum service spacing and the grid VRS minimum service spacing are respectively the service distance with the best positioning effect in the area and the service distance with the worst positioning effect in the area.

[0028] In one possible implementation manner, in the VRS grid division method provided in the embodiment of the present application, in step S3, the grid spacing is set to a value of 3-10 in N types of intervals.

[0029] Beneficial effects: The VRS grid division method of this application takes into account the influence of multiple factors such as the maximum height difference within the grid, the user's spatiotemporal characteristics, and the effective service spacing of the grid, and considers the influence of different factors on the grid division effect for weighted calculation, and finally performs cluster analysis on the calculated grid spacing to optimize the irregular grid coding. Compared with the existing methods, its division is more accurate and detailed, better balancing the VRS calculation pressure on the server side and the positioning effect of the mobile station, greatly improving the positioning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0031] Figure 1 This is the technical flow chart of the VRS grid division method of this application DETAILED DESCRIPTION

[0032] The terms used in the implementation section of this application are only used to explain the specific embodiments of this application and are not intended to limit this application.

[0033] (1) GNSS

[0034] The full name of GNSS is Global Navigation Satellite System, which refers to all satellite navigation systems, including global, regional and enhanced, such as the United States' GPS, Russia's Glonass, Europe's Galileo, China's Beidou satellite navigation system, and related enhancement systems.

[0035] (2)CORS

[0036] CORS stands for Continuously Operating Reference Stations. It establishes multiple continuously operating reference stations at known points on the ground to continuously observe GNSS satellite signals in real time and transmit them to the data processing center for error model solution. The error correction information is then provided to user positioning terminals. While receiving GNSS satellite signals, user positioning terminals also receive error correction information broadcast by the CORS data center and perform differential positioning.

[0037] (3) Base station

[0038] The base station is an important part of the CORS network. It is a device established at a known point on the ground to continuously observe GNSS satellite signals.

[0039] (4) Differential positioning

[0040] Differential positioning means that the positioning terminal receiver receives differential data from the CORS data center to offset errors while receiving GNSS satellite observation data, thereby reducing errors and achieving high-precision positioning. Differential positioning based on CORS mainly includes network real-time code phase differential positioning (Network Real Time Differentia, Network RTD) with decimeter-level accuracy and network real-time carrier phase differential positioning (Network Real Time Kinematic, Network RTK) with centimeter-level accuracy. Among the network RTK technologies, Virtual Reference Station (VRS) technology is the most widely used.

[0041] (5) VRS technology

[0042] In network RTK positioning technology, virtual reference station (VRS) technology is the most widely used. Its principle is to obtain the approximate position of the user's mobile station, interpolate the atmospheric errors of the CORS base stations around the mobile station and generate VRS virtual observation values, and then transmit the VRS to the user. The user's mobile station uses VRS for differential positioning to obtain high-precision positioning results.

[0043] (6) Grid VRS technology

[0044] VRS technology requires that a VRS be assigned to each user mobile station. The VRS calculation pressure on the server side increases linearly with the increase in the number of users, and cannot meet the concurrent application scenarios of massive users. In this context, grid-based VRS technology came into being. The server side no longer calculates VRS based on the approximate position of the user mobile station, but broadcasts it to the outside by prefabricating a certain number of grid VRS. The mobile station determines which grid it is in based on its own approximate position, and then calls the grid VRS data for differential positioning. With grid-based VRS technology, the calculation pressure on the server side is only related to the number of grids, and is no longer related to the number of user mobile stations, which can meet the concurrent service needs of massive users.

[0045] Location services are used to solve the three-dimensional spatial position problem of a person or object at a certain moment, and are mainly implemented by the Global Navigation Satellite System (GNSS). However, the positioning accuracy of directly receiving GNSS satellite signals for single-point positioning is about 5-10m. In areas facing higher-precision positioning needs, such as natural resource survey and monitoring (sub-meter level), precision agriculture (centimeter level), and intelligent driving (centimeter level), it is mainly achieved by establishing a satellite navigation positioning base station network (Continuously Operating Reference Stations, CORS) on the ground to provide network RTK differential positioning. Among network RTK positioning technologies, virtual reference station (VRS) technology is the most widely used. Its principle is to obtain the approximate position of the user's mobile station, perform atmospheric error interpolation modeling on the CORS base stations around the mobile station, and generate VRS virtual observation values, and transmit VRS to the user. The user's mobile station uses VRS for differential positioning to obtain high-precision positioning results.

[0046] VRS technology requires that a VRS be assigned to each user mobile station. The VRS calculation pressure on the server side increases linearly with the increase in the number of users, and cannot meet the concurrent application scenarios of massive users. In this context, grid-based VRS technology came into being. The server side does not calculate VRS based on the approximate position of the user mobile station, but broadcasts it to the outside by prefabricating a certain number of grid VRS. The mobile station determines which grid it is in based on its own approximate position, and then calls the grid VRS data for differential positioning. With grid-based VRS technology, the calculation pressure on the server side is only related to the number of grids, and is no longer related to the number of user mobile stations, which can meet the concurrent service needs of massive users.

[0047] Since the positioning error has spatial correlation. For example, the closer the mobile station is to the VRS station, the higher the positioning accuracy of the mobile station. Therefore, compared with the ordinary VRS technology, the grid-based VRS technology reduces the computing pressure on the server by sacrificing a certain positioning accuracy. The denser the grid, the higher the positioning accuracy, but the computing pressure on the server is greater. On the contrary, the sparser the grid, the lower the positioning accuracy, and the computing pressure on the server is less. Therefore, the VRS grid division method directly affects the user's positioning accuracy and the computing pressure on the server, and is a key issue that needs to be solved by the grid-based VRS technology.

[0048] At present, mainstream CORS data processing software sets the virtual grid spacing through empirical values, which cannot achieve the optimal balance between user positioning accuracy and server-side data solution pressure. In addition, due to differences in geographical location, observation period, terrain environment, and atmospheric delay, there are differences in services across grids.

[0049] Li Ying et al. analyzed the error sources of grid VRS, verified the interpolation effect of the interpolation model with additional elevation correction, and studied the appropriate density of grid division using the ground-based augmentation system in Henan Province as an example. Chen Chong et al. proposed a method to generate irregular grids to fit the terrain. According to the digital elevation model information in the region, an irregular grid that matches the terrain in the region is interpolated. Huang Dingfa et al. proposed a gridded VRS method for wide-area concurrent user services, and studied the effective service spacing of gridded VRS under different terrain conditions. Shi Mingxing proposed a gridded VRS division method based on the analysis of the spatiotemporal characteristics of atmospheric delay, and designed a method for determining the virtual grid spacing and dynamically evaluating service accuracy.

[0050] However, the VRS grid division methods of the above scholars only take into account the terrain, or take into account a few factors such as terrain and atmospheric delay, and directly delineate the grid in the region. The grid pre-division is still completed by the empirical model, which has the following three defects:

[0051] 1. The influence of different factors on the grid division effect is not considered. The grid VRS positioning effect is affected by multiple factors such as geographical location, observation period, terrain environment, atmospheric delay, etc., and the influence of different factors varies. The existing method does not consider the influence of different factors on the grid division effect, but only analyzes a single influencing factor. The grid division effect has a lot of room for improvement.

[0052] 2. The actual spatiotemporal distribution characteristics of user mobile stations and the real-time positioning effect are not considered. The existing grid VRS division methods only perform error analysis from a mathematical perspective. In fact, they can be further optimized according to the spatiotemporal characteristics of user mobile stations. For example, the spatiotemporal availability of users should be negatively correlated with the grid density, that is, the poorer the spatiotemporal availability, the denser the grid should be.

[0053] 3. No coding optimization for irregular grids under complex terrain conditions. The existing grid VRS division method divides the grid into irregular grids after taking into account the terrain and atmospheric delay. In fact, under complex terrain conditions, irregular grid coding will make grid matching of user mobile stations more complicated, affecting the matching efficiency.

[0054] In order to solve the above technical problems, the technical idea of ​​the present application is to propose a VRS grid division method. First, the regional atmospheric delay gradient is calculated according to the historical data of each CORS base station, and the maximum service spacing and the minimum service spacing of the grid VRS in the region are determined according to the set error threshold, and the service spacing is divided into N intervals in an arithmetic progression (N can be determined according to the actual division fineness); then the grid spacing of the N / 2 interval is rounded and the grid is divided into equally spaced grids with longitude and latitude grids; then the grid spacing scaling factor is calculated according to the three factors of the maximum height difference in the grid, the user's spatiotemporal characteristics, and the effective service spacing of the grid (taking into account the influence of different factors on the grid division effect), and the grid spacing is recalculated according to the scaling factor; then the recalculated grid spacing is matched with N grid intervals, and the empirical clustering threshold is selected to perform cluster analysis on the grids with matching interval numbers; finally, the clustered grids are encoded, and the grid relationships at the edges between different clustering areas are processed to ensure that the effective service range of the grid VRS can achieve full regional coverage.

[0055] See also Figure 1 As shown, the VRS grid division method specifically includes the following steps:

[0056] Step S1, calculate the regional atmospheric delay gradient based on the historical observation data of the CORS reference station. The atmospheric delay calculation can be performed by calculating the double difference tropospheric delay and the double difference ionospheric delay, and converting them into error gradients in the N and E directions (i.e., the atmospheric error value corresponding to 1 km). This algorithm is a public method and will not be elaborated in detail in this invention.

[0057] Step S2: Calculate the maximum and minimum grid VRS service spacing that can be set in the region based on the atmospheric delay gradient. Set the positioning error threshold, and calculate the values ​​of the worst positioning effect (VRS minimum service spacing) and the best positioning effect (VRS maximum service spacing) in the region according to the regional atmospheric delay gradient calculated in step S1.

[0058] Step S3: According to the grid VRS maximum service spacing and the grid VRS minimum service spacing calculated in step S2, the grid spacing is set to N types of intervals in the form of an arithmetic progression, and the value of N is set according to the grid division fineness.

[0059] Among them, the grid intervals are generally divided according to the required precision. The finer the division, the more difficult it is to cluster and encode the grids; the coarser the division, the more difficult it is to achieve the optimal balance between the positioning effect and the number of grids. It is generally recommended that the value of N be 3-10.

[0060] Step S4, dividing the area into equal intervals according to longitude and latitude grids using the grid spacing corresponding to rounding N / 2.

[0061] Step S5: Calculate the atmospheric delay and effective service spacing D in each latitude and longitude grid according to the regional atmospheric delay gradient calculated in step S1 and the positioning error threshold set in step S2. i .

[0062] Step S6: Sampling is performed at a spacing of 1 km × 1 km, sampling is performed by fitting the digital elevation model data in the area, and the maximum height difference ΔH in the grid is calculated. i , and set the height difference threshold H, and calculate the height difference influence factor as follows:

[0063]

[0064] Step S7: Calculate the spatiotemporal characteristic index of users in the grid using historical user GGA data. The calculation formula is as follows:

[0065]

[0066] In the formula, UI i is the user spatiotemporal characteristic index of grid i, P USI is the user spatial activity weight, which can be adjusted according to experience or actual test data, n is the number of all users in grid i during the statistical time, N is the total number of grids, and m is the number of users in all grids in the region during the statistical time; P UAI is the user's spatiotemporal availability weight, which can be adjusted based on experience or actual test data. t is the sampling time, which is 1-s and is expressed in days. t is the number of fixed users in grid i on the sampling day, n t is the number of all users in grid i on the sampling day.

[0067] Step S8: Maximum height difference EH in the grid i , User spatiotemporal characteristic index UI i , grid effective service spacing D i The three factors are weighted to calculate the grid spacing scaling factor, where the scaling factors in the N direction and the E direction can be calculated separately. The calculation formula is as follows:

[0068]

[0069] In the formula, SF iis the scaling factor, D is the latitude and longitude grid spacing, P D is the effective service distance weight, P UI is the user's spatiotemporal characteristic weight, P Hi It is the maximum height difference weight within the grid, which can be adjusted based on experience or actual test data.

[0070] Step S9: Divide the latitude and longitude grid spacing D by the scaling factor SF i , and obtain the final spacing of each grid.

[0071] Step S10: Based on the final spacing of each grid in step S9 and the N-type grid spacing division result in step S3, all grids are classified, and the DBSCAN algorithm is used to select the empirical clustering threshold to divide the entire area into N grid clustering spaces.

[0072] Each grid will have an effective spacing value. According to the value space corresponding to the previous step S3, each grid can be defined as an interval. For example, the 1st interval is 4km-5km, and the 2nd interval is 5km-6km. If the calculated grid A is 4.2km, it is defined as a 1st interval. After defining intervals for each grid, each grid will be messy. For example, A is interval 1, and some of the front, back, left and right are 1 and some are 2. Then the irregular grid will cause trouble for encoding, so clustering is required. Clustering is to group those with similar values ​​together and give them the same value. For example, there is only one interval around A that is not 1. After clustering, the one that is not interval 1 will also be assigned interval 1. The threshold parameters of clustering also need to be adjusted repeatedly. The previous classification space is different, and the K value of clustering will also be different, so it needs to be debugged repeatedly.

[0073] Step S11, encode the longitude and latitude of the clustered grids, and check whether there are service gaps in the grid VRS at the edges of different cluster spaces. If there are service gaps, add grid VRS at the edges to ensure full service coverage.

[0074] The VRS grid division method of this application takes into account the influence of multiple factors such as the maximum height difference within the grid, the user's spatiotemporal characteristics, and the effective service spacing of the grid, and considers the influence of different factors on the grid division effect for weighted calculation, and finally performs cluster analysis on the calculated grid spacing to optimize the irregular grid coding. Compared with the existing methods, its division is more accurate and detailed, better balancing the VR S calculation pressure on the server side and the positioning effect of the mobile station, greatly improving the positioning effect.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A VRS grid division method, characterized in that: The steps include: Step S1, calculating the regional atmospheric delay gradient based on the historical observation data of the CORS reference station, and converting it into error gradients in the N and E directions; Step S2, setting a positioning error threshold, and calculating the maximum grid VRS service spacing and the minimum grid VRS service spacing that can be set in the area based on the atmospheric delay gradient; Step S3, according to the grid VRS maximum service spacing and the grid VRS minimum service spacing calculated in step S2, the grid spacing is set to N types of intervals in an arithmetic progression; Step S4, dividing the area into equal intervals according to longitude and latitude grids using the grid spacing corresponding to rounding N / 2; Step S5: Calculate the atmospheric delay and effective service spacing D in each latitude and longitude grid according to the regional atmospheric delay gradient calculated in step S1 and the positioning error threshold set in step S2. i ; Step S6: Sampling is performed at a spacing of 1 km × 1 km, sampling is performed by fitting the digital elevation model data in the area, and the maximum height difference ΔH in the grid is calculated. i , and set the height difference threshold H, and calculate the height difference influence factor as follows: Step S7: Calculate the spatiotemporal characteristic index of users in the grid using historical user GGA data. The calculation formula is as follows: In the formula, UI i is the user spatiotemporal characteristic index of grid i, P USI is the user spatial activity weight, n is the number of all users in grid i during the statistical time, N is the total number of grids, and m is the number of users in all grids in the region during the statistical time; P UAI is the user's spatiotemporal availability weight, t is the sampling time, the value is 1-s, the unit is day, j t is the number of fixed users in grid i on the sampling day, n t is the number of all users in grid i on the sampling day; Step S8: Maximum height difference EH in the grid i , User spatiotemporal characteristic index UI i , grid effective service spacing D i The three factors are weighted to calculate the grid spacing scaling factor, where the scaling factors in the N direction and the E direction can be calculated separately. The calculation formula is as follows: Where SF i is the scaling factor, D is the latitude and longitude grid spacing, P D is the effective service distance weight, P UI is the user's spatiotemporal feature weight, is the maximum height difference weight within the grid; Step S9: Divide the latitude and longitude grid spacing D by the scaling factor SF i , get the final spacing of each grid; Step S10: combining the final spacing of each grid in step S9 and the N-type grid spacing division result in step S3, classifying all grids, using the DBSCAN algorithm, selecting the empirical clustering threshold, and dividing the entire area into N grid clustering spaces; Step S11, encode the longitude and latitude of the clustered grids, and check whether there are service gaps in the grid VRS at the edges of different cluster spaces. If there are service gaps, add grid VRS at the edges to ensure full service coverage.

2. The VRS grid division method according to claim 1, characterized in that: In step S2, the positioning error threshold is that it cannot exceed 2 cm in both the N and E directions.

3. The VRS grid division method according to claim 1, characterized in that: In step S2, the grid VRS maximum service distance and the grid VRS minimum service distance are respectively the service distance with the best positioning effect in the area and the service distance with the worst positioning effect in the area.

4. The VRS grid division method according to claim 1, characterized in that: In step S3, the grid spacing is set to a value of 3-10 in N-type intervals.

Images