A GNSS positioning correction method taking into account diffracted non-line-of-sight signals

Abstract

The present invention discloses a GNSS positioning correction method that takes into account diffracted non-line-of-sight signals. The method is characterized by differentiating NLOS signals into two different types: reflection and diffraction, and using different models for correction in GNSS positioning. Specifically, the method includes: detecting the reflection and diffraction of the NLOS signal; if it is determined to be a reflection signal, calculating the reflection signal delay; if it is determined to be a diffraction signal, calculating the diffraction signal delay. By distinguishing and correcting different NLOS signals, positioning accuracy in urban environments can be effectively improved. Compared with the existing technology, the present invention has the characteristics of dividing NLOS signals into reflection and diffraction signals and correcting them separately, thereby improving positioning accuracy and further optimizing the NLOS signal correction method. This will further enhance the robustness of positioning in urban canyon environments and has better application scenarios.

Description

Technical Field

[0001] The present invention relates to the field of GNSS positioning technology, and in particular to a GNSS positioning correction method taking into account diffracted non-line-of-sight signals. Background Art

[0002] Using GNSS for positioning in urban areas still presents numerous challenges. Nearby obstacles can block the signal's linear propagation, reducing its availability or even rendering it unusable. Multipath interference and NLOS signals are the primary sources of positioning error in urban environments, with NLOS signals causing even greater errors, potentially exceeding 50 meters. Effectively processing NLOS signals is a pressing issue for positioning in urban environments.

[0003] Correctly classifying and identifying signal types is a prerequisite for effective signal processing. Current signal classification methods categorize GNSS signals in urban environments into three types: line-of-sight signals, NLOS signals, and multipath interference signals. These methods do not further distinguish between the different types of NLOS signals, but instead treat them uniformly as reflected NLOS signals. However, the mechanisms and magnitudes of the effects of reflection and diffraction on GNSS signals vary significantly. For example, NLOS reflected signals can cause tens of meters of signal delay in pseudorange, while diffracted signals only cause sub-meter errors. Using a unified approach would introduce new errors into positioning results.

[0004] Existing technologies treat NLOS signals as reflections, which can lead to incorrect estimation of the signal delay due to diffraction, and thus incorrect corrections to observations, affecting positioning accuracy. To date, there have been no publicly reported technologies that distinguish NLOS signals into two distinct types: reflections and diffraction, and apply different correction models to GNSS positioning. Summary of the Invention

[0005] The purpose of the present invention is to provide a GNSS positioning correction method that takes into account diffracted non-line-of-sight signals in response to the shortcomings of the existing technology. The NLOS signals are divided into two different types of signals, reflection and diffraction, and different models are used for correction in GNSS positioning, which effectively improves the positioning accuracy in urban environments. The method is simple and comprehensively considers the propagation characteristics of different types of signals. By distinguishing different NLOS signals, the correction effect is better and the positioning accuracy is higher, which further improves the robustness of positioning in urban canyon environments and has good application scenarios.

[0006] The specific technical solution for achieving the purpose of the present invention is: a GNSS positioning correction method that takes into account diffracted non-line-of-sight signals. The method is characterized by considering that NLOS signals propagated by reflection and diffraction exhibit different error characteristics, distinguishing NLOS signals into two different types of signals, namely reflected and diffracted signals, and using different models for correction in GNSS positioning. The method specifically includes the following steps:

[0007] Step 1: Use the 3D urban environment model to obtain the distances w and w′ from the observation point to the buildings on both sides, as well as the heights h and h′ of the buildings on both sides. Compare the satellite elevation angle with the elevation angle of the buildings at the same azimuth to determine whether the satellite signal is an NLOS signal.

[0008] Step 2: If an NLOS signal is detected, the elevation angle of the first Fresnel zone boundary is calculated using geometric relationships based on the satellite's azimuth and elevation angles and the radius of the first Fresnel zone. If the elevation angle of the first Fresnel zone boundary is less than the elevation angle of the building at the corresponding azimuth, the first Fresnel zone is considered completely obscured.

[0009] Step 3: If the first Fresnel zone of the signal is not completely blocked, determine whether there is a diffraction point. If there is a diffraction point, the signal is determined to be a diffraction signal, and the diffraction correction value is calculated using the UTD theory.

[0010] Step 4: If the first Fresnel zone of the signal is completely blocked, determine whether there is a reflection point. If there is a reflection point, determine that the signal is a reflection signal and calculate the reflection correction value.

[0011] Step 5: Compensate the original observations with the diffraction correction Δd and reflection correction Δb calculated in steps 3 and 4, correct the signal delay caused by the environment, and finally use the corrected observations for positioning.

[0012] In step 1, the coordinate relationship between the observation point coordinates and the key corner points of the building can be used to calculate the distance between the observation point and the buildings on both sides and the height of the buildings on both sides, and finally compare the satellite elevation angle and the elevation angle of the building at the same azimuth.

[0013] In step 2, the radius of the first Fresnel zone is calculated by the following formula (a):

[0014]

[0015] Among them, R F is the radius of the first Fresnel zone at a specified point during signal propagation; λ is the carrier wavelength; d s is the distance between a specified point and the satellite; d A is the distance between a specified point and the observation point; d is the distance between the satellite and the observation point.

[0016] In step 2, the elevation angle of the first Fresnel zone boundary is obtained by the following formula (b):

[0017]

[0018] Among them, el j is the elevation angle corresponding to the upper boundary point of the first Fresnel zone radius; el is the satellite elevation angle; Az is the satellite azimuth; az j is the azimuth angle corresponding to the upper boundary point of the first Fresnel zone radius.

[0019] The azimuth angle az corresponding to the upper boundary point of the first Fresnel zone radius j It represents the integer azimuth values contained in the azimuth range of the first Fresnel zone, where the azimuth set of the first Fresnel zone is calculated by the following formula (c):

[0020]

[0021] Among them, [] is the rounding symbol, az j Get the integer value from the set az.

[0022] The diffraction signal determination in step 3 is based on the diffraction angle β0 of the satellite signal during diffraction propagation and the minimum diffraction angle β that can occur on the building boundary. min and the maximum diffraction angle β max Make a judgment, when β min <β0<β max , then it is judged that there is a diffraction point, and the diffraction correction value Δd is calculated using the UTD theory.

[0023] The diffraction angle β0 is calculated by the following formula (d):

[0024] β0=acos(sin(Az)*cos(El)) (d).

[0025] Among them, Az is the azimuth angle of the satellite; El is the altitude angle of the satellite.

[0026] The minimum diffraction angle β min Calculated by the following formula (e):

[0027]

[0028] Among them, x min is the minimum coordinate value on the boundary of the building on the side blocking the signal; h is the height of the building on the side blocking the signal; w is the distance from the building on the side blocking the signal.

[0029] The maximum diffraction angle β max Calculated by the following formula (f):

[0030]

[0031] Among them, x max It is the maximum coordinate value on the building boundary on the side blocking the signal.

[0032] The diffraction correction value Δd is obtained by the following formula (g):

[0033]

[0034] Among them, x Qe is the coordinate value of the diffraction point on the building boundary on the side blocking the signal; El is the altitude angle of the satellite; Az is the azimuth angle of the satellite.

[0035] The reflection point in step 4 is determined by the following formula (h):

[0036]

[0037] If equation (g) above holds true, it can be determined that there is a signal reflection point on the building, and the reflection correction value Δb can be calculated using equation (i) below:

[0038] Δb=2*w*cos(Az)*cos(El) (i).

[0039] Compared with the existing technology, the present invention has better correction effect and higher positioning accuracy. It distinguishes NLOS signals into two different types of signals, reflection and diffraction, and adopts different models for correction in GNSS positioning, which can effectively ensure the accuracy of signal delay correction and effectively improve positioning accuracy. This will further enhance the robustness of positioning in urban environments and has better application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 Flowchart of the present invention;

[0041] Figure 2 This is the NLOS judgment result diagram of the C28 satellite;

[0042] Figure 3 This is the NLOS judgment result diagram of the G31 satellite;

[0043] Figure 4 This is a dual-axis diagram for the pseudorange residuals and NLOS diffraction signal judgment of the C28 satellite;

[0044] Figure 5 This is a dual-axis diagram for the pseudorange residuals and NLOS diffraction signal judgment of the G31 satellite;

[0045] Figure 6This is a dual-axis diagram for the pseudorange residuals and NLOS reflection signal judgment of the C28 satellite;

[0046] Figure 7 This is a dual-axis diagram for the pseudorange residuals and NLOS reflection signal judgment of the G31 satellite;

[0047] Figure 8 This is the dual-axis diagram of pseudorange correction value and pseudorange residual of G31 satellite;

[0048] Figure 9 Comparison of the root mean square error of the positioning correction method using the diffraction NLOS signal and the one not taking the diffraction NLOS signal into account in the east, north and sky directions. DETAILED DESCRIPTION

[0049] See Figure 1 , the present invention specifically includes the following steps:

[0050] Step 1: Use the coordinate relationship between the observation point coordinates and the key corner points of the building to calculate the distance to the buildings on both sides and the height of each building on both sides. Compare the satellite's elevation angle with the elevation angle of the building at the corresponding azimuth angle to determine whether it is an NLOS signal.

[0051] Step 2: If an NLOS signal is detected, the elevation angle of the first Fresnel zone boundary is calculated using geometric relationships based on the satellite azimuth, elevation angle, and radius of the first Fresnel zone. If the elevation angle values of the first Fresnel zone boundary are all less than the elevation angle values of the building at the corresponding azimuth, the first Fresnel zone is determined to be completely blocked.

[0052] Step 3: If the first Fresnel zone of the signal is not completely blocked, the diffraction angle can be calculated using the following formula (d). The minimum and maximum diffraction angles that can occur on the building boundary can be calculated using formulas (e) and (f). When β min <β0<β max , then it is judged that there is a diffraction point, and the correction value can be calculated using formula (g).

[0053] β0=acos(sin(Az)*cos(El)) (d).

[0054] Among them, β0 is the diffraction angle of the satellite signal during diffraction propagation, Az is the azimuth angle of the satellite, and El is the altitude angle of the satellite.

[0055]

[0056]

[0057] Among them, β min and β maxare the minimum and maximum diffraction angles at which diffraction can occur on the building boundary, x min is the minimum coordinate value on the building boundary on the side blocking the signal, x max is the maximum coordinate value on the boundary of the building on the side of the blocked signal, h is the height of the building on the side of the blocked signal, and w is the distance from the building on the side of the blocked signal.

[0058]

[0059]

[0060] Where Δd is the diffraction calculation correction value, x Qe is the coordinate value of the diffraction point on the boundary of the building on the side blocking the signal, h is the height of the building on the side blocking the signal, w is the distance from the building on the side blocking the signal, El is the altitude angle of the satellite, and Az is the azimuth angle of the satellite.

[0061] Step 4: Determine whether there is a signal reflection point on the building. If formula (h) holds true, it can be determined that there is a signal reflection point on the building, and the correction value can be calculated according to formula (i).

[0062]

[0063] Where h is the building height, w is the distance between the observation point and the building, Az is the satellite azimuth, and El is the satellite elevation angle.

[0064] Δb=2*w*cos(Az)*cos(El) (i).

[0065] Among them, Δb is the reflection calculation correction value, w is the distance between the observation point and the building, Az is the satellite azimuth, and El is the satellite altitude angle.

[0066] Step 5: Compensate the original observations with the signal correction values calculated using formulas (g) and (i) in the previous steps to correct the signal delay caused by the environment. Finally, use the corrected observations for positioning.

[0067] The present invention is further described below through a specific embodiment of an experimental environment in a certain campus.

[0068] Example 1

[0069] In step 1, a ground-based laser radar is used to obtain point cloud data of buildings around an urban canyon environment on a certain campus. The coordinate relationship between the observation point coordinates and the coordinates of the key corner points of the buildings is used to calculate the distances w = 18.476m, w′ = 15.715m to the buildings on both sides and the heights h = 22.075m and h′ = 20.572m of the buildings on both sides. The elevation angle of the satellite and the elevation angle of the building at the corresponding azimuth angle are compared to determine whether it is an NLOS signal.

[0070] See Figure 2 , the C28 satellite was observed to be an NLOS signal during the entire observation period from epoch 10961 to 13544.

[0071] See Figure 3 The G31 satellite was observed to be an NLOS signal during the entire observation period from epoch 6735 to 10287.

[0072] In step 2, if an NLOS signal is detected, the elevation angle information of the first Fresnel zone boundary is calculated using geometric relationships in combination with the satellite azimuth, elevation angle, and radius of the first Fresnel zone. That is, the zone is divided according to a resolution of 1° in azimuth, and the elevation angle value of the first Fresnel zone boundary at the corresponding azimuth angle is calculated. If the elevation angle values of the first Fresnel zone boundary are all less than the elevation angle value of the building at the corresponding azimuth angle, it is determined that the first Fresnel zone is completely blocked.

[0073] In step 3, if the first Fresnel zone of the signal is not completely blocked, the diffraction angle can be calculated using formula (c). The minimum and maximum diffraction angles that can occur on the building boundary can be calculated using formulas (d) and (e). min <β0<β max , then it is judged that there is a diffraction point, and the correction value can be calculated using formula (h).

[0074] See Figure 4 The C28 satellite was judged to be an NLOS diffraction signal during the entire observation period from epoch 12338 to 13544. At the same time, the variation pattern of the pseudorange residual of the satellite also conforms to the error characteristics of the diffraction signal.

[0075] See Figure 5 The G31 satellite was determined to be an NLOS diffraction signal during the entire observation period from epoch 6735 to 7646. At the same time, the variation pattern of the pseudorange residuals of this satellite also conforms to the error characteristics of the diffraction signal.

[0076] In step 4, it is determined whether there is a signal reflection point on the building. If formula (g) holds true, it can be determined that there is a signal reflection point on the building, and the correction value can be calculated according to formula (h).

[0077] See Figure 6 The C28 satellite was judged to be an NLOS reflection signal during the entire observation period from epoch 10961 to 12338. During this period, the satellite's pseudorange error showed a large positive error, which is consistent with the characteristics of the extra optical path when the signal is reflected.

[0078] See Figure 7 The G31 satellite was observed to be an NLOS reflection signal from epoch 7647 to 10287. During this period, the satellite's pseudorange error showed a large positive error, which is consistent with the characteristics of the extra optical path when the signal is reflected.

[0079] In step 5, the signal correction values calculated according to formulas (f) and (h) in the previous steps are used to compensate for the original observations, correct the signal delay caused by the environment, and finally use the corrected observation values for positioning solution.

[0080] See Figure 8 The variation patterns of the correction values and pseudorange residuals of the G31 satellite are basically consistent, indicating that the comprehensive correction method can accurately eliminate the additional signal delay part to a certain extent.

[0081] See Figure 9 By comparing the error distribution diagrams of the two within the intercepted observation epoch interval, it can be seen that the correction method taking into account the diffraction NLOS signal is better than the original correction method. The RMSE in the east direction is improved from 13.539m to 7.801m, the RMSE in the north direction is improved from 15.421m to 13.181m, and the RMSE in the sky direction is improved from 30.631m to 22.510m.

[0082] The above is only a further explanation of the present invention and is not intended to limit this patent. Any equivalent implementation of the present invention should be included in the scope of the claims of this patent.

Claims

1. A GNSS positioning correction method taking into account diffracted non-line-of-sight signals, characterized in that: NLOS signals are divided into two types: reflection and diffraction. Different correction models are used in GNSS positioning. The specific steps include: Step 1: Use the 3D urban environment model to obtain the distances w and w′ from the observation point to the buildings on both sides, as well as the heights h and h′ of the buildings on both sides. Compare the satellite elevation angle el with the elevation angles of the buildings at the same azimuth to determine whether the satellite signal is an NLOS signal. Step 2: If the signal detected is an NLOS signal, then combine the satellite azimuth angle Az, the satellite elevation angle el and the first Fresnel zone radius R at a specified point during the signal propagation process F , get the altitude angle az corresponding to the upper boundary of the first Fresnel zone radius j , if the altitude angle value az j If the building elevation angle is smaller than the corresponding azimuth angle, the first Fresnel zone is considered to be completely blocked; Step 3: If the first Fresnel zone of the signal is not completely blocked, determine whether there is a diffraction point. If there is a diffraction point, determine that the signal is a diffraction signal and calculate the diffraction correction value Δd; Step 4: If the first Fresnel zone of the signal is completely blocked, determine whether there is a reflection point. If there is a reflection point, determine that the signal is a reflection signal and calculate the reflection correction value Δb; Step 5: Compensate the original observations based on the calculated diffraction correction value Δd and reflection correction value Δ to correct the signal delay caused by the environment, and use the corrected observations for positioning. The step 3 is based on the diffraction angle β0 of the satellite signal during diffraction propagation and the minimum diffraction angle β that can occur on the building boundary. min and the maximum diffraction angle β max To determine the diffraction point, when β min <β0<β max , then it is determined that there is a diffraction point, and the diffraction correction value Δd is calculated using the UTD theory; The diffraction angle β0 is calculated by the following formula (d): β0=acos(sin(Az)*cos(El)) (d); Where Az is the azimuth angle of the satellite; El is the altitude angle of the satellite; The minimum diffraction angle β min Calculated by the following formula (e): Among them, x min is the minimum coordinate value on the boundary of the building on the side blocking the signal; h is the height of the building on the side blocking the signal; w is the distance from the building on the side blocking the signal; The maximum diffraction angle β max Calculated by the following formula (f): Among them, x max is the maximum coordinate value on the building boundary on the side blocking the signal; The reflection point in step 4 is determined by the following formula (h):

2. The GNSS positioning correction method taking into account diffracted non-line-of-sight signals according to claim 1, characterized in that: The radius R of the first Fresnel zone in step 2 F Calculated by the following formula (a): Where λ is the carrier wavelength; d s is the distance between a specified point and the satellite; d A is the distance between a specified point and the observation point; d is the distance between the satellite and the observation point; The elevation angle el corresponding to the upper boundary of the radius of the first Fresnel zone j Calculated by the following formula (b): Among them, el is the satellite elevation angle; Az is the satellite azimuth angle; az j is the azimuth angle corresponding to the upper boundary point of the first Fresnel zone radius, which represents the integer azimuth angle value contained in the azimuth angle range of the first Fresnel zone. The azimuth angle set of the first Fresnel zone is calculated by the following formula (c): Among them, [] is the rounding symbol; az j Get the integer value from the set az.

3. The GNSS positioning correction method taking into account diffracted non-line-of-sight signals according to claim 1, characterized in that: The diffraction correction value Δd is calculated by the following formula (g): Among them, x Qe is the coordinate value of the diffraction point on the building boundary on the side blocking the signal; El is the altitude angle of the satellite; Az is the azimuth angle of the satellite.

4. The GNSS positioning correction method taking into account diffracted non-line-of-sight signals according to claim 1, characterized in that: If equation (h) in step 4 holds true, it can be determined that there is a signal reflection point on the building, and the reflection correction value Δb can be calculated using equation (i) below: Δb=2*w*cos(Az)*cos(El) (i).

Images