Three-dimensional laser radar assisted high-precision satellite positioning method

Abstract

The invention provides a method for supporting vehicle positioning by using a satellite positioning system. The method includes: generating a sliding window map (SWM) in real-time based on a 3D point cloud from a 3DLiDAR sensor and an attitude heading reference system (AHRS), where the SWM provides an environmental description for detection and NLOS reception; accumulating a 3D point cloud from a previous frame into the SWM to enhance the field of view of the 3D LiDAR sensor; receiving, by the GNSS receiver, GNSS measurements from the satellite; detecting the NLOS reception from the GNSS measurements using the SWM; when a reflection point is not found in the SWM, correcting the NLOS reception by NLOS reconstruction; and estimating GNSS positioning through a least square algorithm. An object of the present application is to provide a method of mitigating NLOS reception caused by static buildings and dynamic objects.

Description

Technical Field

[0001] The present disclosure relates generally to the field of autonomous driving or other types of autonomous systems for intelligent transportation systems. Specifically, the present disclosure relates to a 3D LiDAR-assisted global navigation satellite system and method for NLOS detection and correction, which can improve positioning performance. Background Art

[0002] Autonomous driving is widely recognized as a remedy for excessive traffic congestion and anticipated accidents. However, the insufficient positioning accuracy of current solutions is one of the key issues hindering the advent of autonomous driving in urban scenarios. As the demand for ADVs continues to increase, positioning in urban environments has become crucial.

[0003] GNSS, such as GPS and / or other similar satellite-based positioning technologies, is currently one of the primary means of providing global reference positioning for ADV positioning in intelligent transportation systems. With the increasing availability of multiple satellite constellations, GNSS can provide satisfactory performance in open sky areas. However, when large parts of the sky are obscured, GNSS performance degrades significantly, which is a challenging problem. This situation is known as the "urban canyon" scenario. Typically, in highly urbanized cities, positioning accuracy degrades significantly due to signal reflections caused by static buildings and dynamic objects.

[0004] In particular, GNSS-RTK is used for high-precision aerial mapping and positioning of Level 4 fully autonomous vehicles. Generally, GNSS-RTK positioning involves two steps: (1) estimating a floating-point solution based on the received GNSS measurements; (2) resolving integer ambiguities using a least squares algorithm (e.g., LAMBDA) based on the derived floating-point solution as an initial guess. When a fixed solution is achieved, centimeter-level positioning accuracy can be achieved based on double differential carrier and coded measurements in open areas. Unfortunately, the accuracy of GNSS-RTK is significantly reduced in urban canyons due to NLOS and multipath reception caused by GNSS signal reflections and obstruction by surrounding buildings. In practice, the significant decrease in GNSS-RTK positioning accuracy in urban canyons is mainly caused by the occurrence of GNSS NLOS reception. Due to NLOS reception, the performance of GNSS positioning is highly affected by real-time surrounding environment features such as buildings and dynamic objects. Some of the received GNSS signals are severely contaminated, including loud noise. According to the inventors' previous research, in highly urbanized areas, most of the received GNSS signals may be multipath or NLOS reception. Therefore, the accuracy of the floating-point solution estimation based on differential carrier and code measurements is reduced, making it difficult to obtain a fixed solution for ambiguity resolution.

[0005] Furthermore, the number of available satellites in urban canyons is limited due to signal obstruction from surrounding buildings. Consequently, the geometry of the satellite distribution is distorted, resulting in a large DOP. Consequently, due to poor satellite geometry, the search space ambiguity is large, making it difficult to obtain a fixed solution. In short, urban canyon scenarios introduce additional difficulties for both steps of GNSS-RTK positioning.

[0006] Therefore, effectively sensing and understanding the surrounding environment is key to improving GNSS positioning in urban areas, as GNSS positioning heavily relies on sky visibility. The best-known approach to addressing GNSS NLOS reception is 3DMA GNSS positioning, such as NLOS exclusion and shadow matching based on 3D map construction information. However, the shortcomings of these 3DMA GNSS methods are: 1) they rely on the availability of 3D building models and an initial guess of the GNSS receiver's position; and 2) they are unable to mitigate NLOS reception caused by surrounding dynamic objects. Recent advances in 3DMA GNSS positioning methods are reviewed in detail in our previous work.

[0007] In a recent publication, the inventors demonstrated the use of 3D LiDAR sensors, often referred to as the "eyes" of autonomous vehicles (ADVs), as a typical and indispensable onboard sensor for autonomous vehicles, to detect NLOS caused by dynamic objects. Due to the 3D LiDAR's limited -30° to +10° field of view (FOV), only a portion of a double-decker bus could be scanned. Furthermore, this method relies heavily on object detection accuracy. Nevertheless, this is the first work to employ real-time object detection to assist GNSS positioning. The authors explored using real-time 3D LiDAR point clouds to detect surrounding buildings, rather than just dynamic objects. Due to the 3D LiDAR's limited field of view, only a portion of buildings could be scanned. Therefore, information about building heights was required to detect NLOS caused by buildings. Rather than eliminating detected NLOS reception, the inventors explored alternative approaches to correct NLOS pseudorange measurements with the help of LiDAR. 3D LiDAR can measure the distance from the GNSS receiver to building surfaces that may have reflected GNSS signals. Both the corrected and remaining healthy GNSS measurements can then be used for further GNSS positioning. Correcting for detected NLOS satellites yielded improved performance. However, the performance of this approach relies on the accuracy of building and reflector detection. Both building and reflector detection may fail when the building surface is highly irregular. The limited FOV of LiDAR remains a disadvantage in the detection of dynamic objects and buildings. Overall, previous work has shown the feasibility of detecting GNSS NLOS using real-time airborne sensing (real-time point cloud). To overcome the disadvantage of the limited FOV of 3D LiDAR, the inventors explored the use of fisheye cameras and 3D LiDAR to detect and correct NLOS signals. The fisheye camera is used to detect non-line-of-sight signals. At the same time, 3D LiDAR is used to measure the distance between the GNSS receiver and the potential reflectors that cause NLOS signals. However, this approach has the problem that NLOS detection is sensitive to ambient lighting conditions.

[0008] Therefore, there is a need in the art for improved positioning methods and systems, particularly for autonomous driving capable of achieving high-precision positioning in very deep urban canyons. Moreover, other desirable features and characteristics will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and background of the present disclosure. Summary of the Invention

[0009] This paper provides a 3D LiDAR-assisted GNSS NLOS mitigation method and a system for implementing the method. The purpose of this disclosure is to provide a method for mitigating NLOS caused by static buildings and dynamic objects.

[0010] A first aspect of the present disclosure provides a method for supporting vehicle positioning using a satellite positioning system. The method includes: receiving LiDAR factors and IMU factors from a 3D LiDAR sensor and a LIO; integrating the LiDAR factors and the IMU factors using a local factor graph optimization to estimate the relative motion between two epochs; generating a 3D PCM as an auxiliary landmark satellite for providing low-elevation auxiliary landmark satellites; receiving GNSS measurements from a satellite via a GNSS receiver; detecting GNSS NLOS reception from the GNSS measurements using the 3D PCM; and excluding the GNSS NLOS reception from the GNSS measurements to obtain surviving GNSS satellite measurements, thereby improving the quality of the GNSS measurements used for positioning the autonomous vehicle.

[0011] In one embodiment, the method further comprises: performing GNSS-RTK float estimation on the surviving GNSS satellite measurements to obtain a float solution; performing ambiguity resolution to obtain a fixed ambiguity solution; and determining a fixed GNSS-RTK positioning solution from the float solution and the fixed ambiguity solution.

[0012] In one embodiment, the ambiguity resolution is performed by applying the LAMBDA algorithm.

[0013] In one embodiment, the method further includes: feeding back the fixed GNSS-RTK positioning solution to the 3D LiDAR sensor and the LIO; and performing PCM correction using the fixed GNSS-RTK positioning solution to correct the drift of the 3D point cloud.

[0014] In one embodiment, the method further comprises: obtaining an initial guess of the float solution using a least squares algorithm on the GNSS measurements.

[0015] A second aspect of the present disclosure provides a method for supporting vehicle positioning using a satellite positioning system. The method includes: generating a sliding window map (SWM) in real time based on a 3D point cloud from a 3D LiDAR sensor and an AHRS, wherein the SWM provides an environment description for detecting and correcting NLOS reception; accumulating 3D point clouds from previous frames into the SWM to enhance the field of view of the 3D LiDAR sensor; receiving GNSS measurements from satellites via a GNSS receiver; detecting the NLOS reception from the GNSS measurements using the SWM; correcting the NLOS reception by NLOS reconstruction when a reflection point is not found in the SWM; and estimating the GNSS position using a least squares algorithm.

[0016] In one embodiment, the method further comprises minimizing the point cloud capacity of the SWM by excluding point clouds far from the GNSS receiver so that the 3D point cloud is within a sliding window.

[0017] In one embodiment, the step of generating the SWM includes: obtaining a local map from LiDAR scan matching based on the 3D point cloud from a 3D LiDAR sensor; and converting the SWM from a carrier coordinate system to a local East-North-Upper (ENU) coordinate system using the direction of an AHRS.

[0018] In one embodiment, the step of detecting the NLOS reception from the GNSS measurements is performed using a fast search method, wherein the fast search method includes: initializing a search point at the center of the 3D LiDAR sensor; determining a search direction connecting the GNSS receiver and the satellite based on the elevation and azimuth angles of the satellite; and determining a search direction at a fixed increment value Δd. pix Move the search point along the search direction; calculate the adjacent points (N k ) quantity; and if N k Exceeds the predetermined threshold N thres , the search point is classified as an NLOS satellite.

[0019] In an embodiment, the method further comprises correcting the NLOS reception by re-estimating the GNSS measurements using a model calibration based on the SWM, wherein the SWM provides a dense, discrete and unorganized 3D point cloud without continuous building surfaces or boundaries.

[0020] In one embodiment, the model calibration includes: using an efficient kdTree structure based on a reflector detection algorithm to detect the reflection point corresponding to the NLOS reception, wherein the reflector detection algorithm includes: traversing all azimuths from 0° to 360° with an azimuth resolution of α res , the elevation angle is When the connection point p j detecting a potential reflector when the line of sight to the satellite is not blocked; and detecting a unique reflector having a shortest distance between the GNSS receiver and the potential reflector.

[0021] In one embodiment, NLOS performs the NLOS reconstruction using a weighting scheme with a scaling factor. The scaling factor is used to de-weight the NLOS reception, and the weighting scheme includes the following definitions: if a satellite is classified as a LOS measurement, the scaling factor is calculated based on a satellite signal-to-noise ratio (SNR) and an elevation angle; if the satellite is classified as a NLOS measurement and a pseudorange error is corrected, the scaling factor is calculated based on the satellite SNR and the elevation angle; and if the satellite is classified as a NLOS measurement but no reflection point is detected, the scaling factor is calculated based on the satellite SNR, the elevation angle, and the scaling factor K. w The scaling factor is calculated.

[0022] This summary is provided to introduce a selection of concepts in a simplified form that are further described in the detailed description below. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to serve as an aid in determining the scope of the claimed subject matter. Additional aspects and advantages of the present invention are disclosed as illustrated in the following examples. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are included to further illustrate and clarify the above and other aspects, advantages, and features of the present disclosure. It should be understood that these drawings depict only certain embodiments of the present disclosure and are not intended to limit its scope. It should also be understood that these drawings are illustrated for simplicity and clarity and are not necessarily drawn to scale. The present disclosure will now be described and explained with additional specificity and detail through the use of the accompanying drawings, in which:

[0024] Figure 1 is a schematic diagram of a 3D LiDAR-assisted GNSS real-time dynamic differential positioning method according to an exemplary embodiment of the present disclosure;

[0025] Figure 2 is a schematic diagram of a 3D LiDAR-assisted GNSS NLOS mitigation method according to an exemplary embodiment of the present disclosure;

[0026] Figure 3 is an exemplary demonstration of a sliding window map and real-time 3D point cloud generated according to an exemplary embodiment of the present disclosure;

[0027] Figure 4 is a schematic diagram of the coordinate system used in the present invention;

[0028] Figure 5 is a conceptual diagram illustrating NLOS detection based on a generated sliding window map according to an exemplary embodiment of the present disclosure;

[0029] Figure 6 It reconstructs and corrects the polluted GNSS signals by sensing the world;

[0030] Figure 7 The positioning performance before and after the application of the 3D LiDAR-assisted GNSS real-time dynamic differential positioning method according to an embodiment of the present invention;

[0031] Figure 8 is a diagram of a 3D LiDAR-assisted GNSS system configured to utilize a 3D LiDAR-assisted GNSS NLOS mitigation method and incorporated into a vehicle according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0032] The present disclosure is primarily directed to the field of autonomous driving or other types of autonomous systems requiring navigation using satellite positioning systems. The vehicle may be an autonomous vehicle (ADV) or a vehicle equipped with ADAS. More specifically, but not limited to, the present disclosure provides a 3D LiDAR-assisted GNSS NLOS mitigation method that can detect and correct NLOS signals in very deep urban canyons, and a system implementing the method. The present disclosure is intended to provide a method for mitigating NLOS caused by static buildings and dynamic objects, thereby achieving high accuracy in highly densely urbanized areas.

[0033] Effects, advantages, solutions to problems, and any element that may cause any effect, advantage, or solution to occur or become apparent should not be construed as a critical, required, or essential feature or element of any or all the claims. The invention is defined solely by the appended claims including any amendments made during the pendency of this application and all equivalents of those claims as issued.

[0034] In the following claims and the foregoing description of the invention, unless the context requires otherwise due to express language or necessary implication, the word "comprise" (or variations thereof such as "comprises" or "comprising") is used in an inclusive manner, i.e., to specify the presence of stated features but not to exclude the presence or addition of further features in various embodiments of the invention.

[0035] As used herein, the term "Global Navigation Satellite System" or "GNSS" refers to a general-purpose satellite-aided navigation system in which electronic receivers can determine their position with a specified accuracy using line-of-sight radio signals transmitted by satellites. A GNSS receiver can receive and process signals from multiple satellites orbiting the Earth to determine the GNSS receiver's position, which is then used to determine the vehicle's position. For the purposes of this disclosure, unless otherwise specified, the status of the GNSS receiver and the positions of the satellites are expressed in the ECEF coordinate system.

[0036] The term "LiDAR sensor" refers to a sensor capable of measuring the distance from a vehicle to an object by emitting a laser beam to the object and measuring the reflected portion of the laser beam to calculate the distance.

[0037] The term "attitude heading reference system" or "AHRS" refers to a system having one or more sensors configured to use vibrations to measure changes in the vehicle's direction, orientation, and / or acceleration based on a vertical reference.

[0038] The first embodiment of this disclosure is to detect and eliminate GNSS NLOS reception to further improve GNSS-RTK positioning. It is also important to use improved GNSS-RTK positioning to correct for 3D point cloud drift, thereby improving overall positioning accuracy.

[0039] Figure 1 This paper provides an overview of the 3D LiDAR-assisted GNSS-RTK positioning method proposed in this disclosure. The system consists of two parts: (1) real-time environment description generation based on the cloud from 3D LiDAR and IMU, and corrections from GNSS-RTK solution S100A; (2) GNSS NLOS detection and elimination based on the real-time environment description, and GNSS-RTK positioning based on surviving satellites in S100B.

[0040] In certain embodiments, the use of a 3D LiDAR sensor 110 improves GNSS-RTK positioning in urban canyons by fundamentally addressing the issues inherent in traditional GNSS-RTK due to signal reflections and occlusions. First, the 3D LiDAR sensor 110 and LIO 120 receive and loosely integrate LiDAR and IMU factors using local factor graph optimization 141 to estimate the relative motion between two epochs and generate a 3D PCM to provide an environment description. The LiDAR factors are derived from LiDAR scan matching 111, and the IMU factors are derived from pre-integration 121. The environment description used here is a local environment description. In certain embodiments, the 3D LiDAR sensor 110 is a Velodyne 32 configured to collect raw 3D point cloud data at a rate of 10 Hz, and the LIO 120 is an Xsens Ti-10 IMU configured to collect data at a rate of 100 Hz. Furthermore, the positions of the surrounding point clouds are estimated simultaneously. PCM correction 142 is then performed using the position estimates of the surrounding point clouds from the 3D PCM. The effect is that this method can advantageously generate locally accurate PCMs.

[0041] Second, potential GNSS NLOS satellites are detected and eliminated with the help of the environmental description. A GNSS receiver 130 receives GNSS measurements from the satellites, which can be used to derive an initial guess for a float solution using a conventional least squares algorithm 131. In certain embodiments, the GNSS receiver 130 is a commercial-grade u-blox F9P GNSS receiver that collects raw GPS / Beidou measurements at a data rate of 1 Hz. Additionally, a NovAtel SPAN-CPT, GNSS (GPS, GLONASS, and Beidou) RTK / INS (fiber optic gyroscope, FOG) integrated navigation system can be used to provide ground-truth positioning. All data is collected and synchronized using ROS.

[0042] For subsequent GNSS NLOS detection, the first embodiment of the present disclosure does not rely on an initial guess of a float solution from the GNSS receiver 130. Instead, GNSS NLOS receptions are detected from GNSS measurements using a 3D PCM generated using LiDAR or an inertial integration and correlation algorithm. Thus, the use of GNSS NLOS rejection 151 can alleviate the problem of poor GNSS measurement quality by essentially eliminating potential GNSS NLOS receptions to obtain surviving GNSS satellite measurements.

[0043] To address this issue, the present disclosure proposes using landmarks from the generated point cloud map as "auxiliary landmark satellites" to fundamentally improve satellite geometry. Advantageously, the auxiliary landmark satellites and the receiving satellites complement each other as the point cloud map, providing low-elevation auxiliary landmark satellites. Such low-elevation auxiliary landmark satellites are typically unavailable for physically received satellites. This eliminates contaminated GNSS NLOS satellites and improves satellite geometry with the help of auxiliary landmark satellites.

[0044] Next, a float solution can be estimated based on the surviving GNSS satellite measurements by performing GNSS-RTK float estimation 152 to obtain a float solution, and then performing ambiguity resolution 153 to obtain a fixed ambiguity solution. In some embodiments, ambiguity resolution 153 is performed by applying the LAMBDA algorithm. Finally, the estimated fixed GNSS-RTK positioning solution 154 is fed back to the 3D LiDAR sensor and LIO to perform PCM correction 142 to further correct the drift of the 3D point cloud. By combining the improved GNSS-RTK positioning assisted by GNSS NLOS detection, this method can essentially correct the drift of the 3D point cloud map generated by LiDAR or inertial integration. Therefore, the proposed method effectively combines the complementary properties of LIO (in a locally accurate manner in a short time and providing environmental description for GNSS NLOS detection) and GNSS-RTK (drift-free, with global reference positioning but affected by GNSS NLOS).

[0045] The first embodiment of the present disclosure addresses this issue by leveraging the sensing capabilities of a 3D LiDAR sensor to detect and eliminate potential GNSS NLOS reception. This approach can achieve an accuracy of 10 cm even in urban canyon scenarios, meeting the navigation requirements of autonomous driving. In the second embodiment of the present disclosure, an alternative method and system are provided that employs 3D LiDAR to assist with GNSS single-point positioning. Positioning is only accurate within a range of 5 meters, which is not as accurate as the GNSS-RTK of the first embodiment.

[0046] refer to Figure 2 , illustrates an overview of the 3D LiDAR-assisted GNSS NLOS mitigation method. The system consists of two parts: (1) real-time SWM generation S200A based on 3D point clouds from 3D LiDAR sensors and AHRS; (2) GNSS NLOS detection and correction S200B based on real-time environment description.

[0047] Advantageously, the SWM is generated based on a real-time 3D point cloud from the 3D LiDAR sensor 110. Only the 3D point cloud within a sliding window is used to generate the SWM, as point clouds far from the GNSS receiver 130 are unnecessary for NLOS detection. Therefore, the size of the SWM can be minimized. The SWM is first generated using the 3D LiDAR sensor 110 and the AHRS 160. The SWM essentially provides a description of the environment for detecting and correcting NLOS reception. Based on the 3D point cloud from the 3D LiDAR sensor 110, a local map is obtained from LiDAR scan matching 211. Directions from the AHRS 160 are directly used to transform the SWM from the vehicle coordinate system to the local ENU coordinate system.

[0048] By performing local transformation 241, a SWM can be generated, which is executed in real time for GNSS NLOS detection. This is crucial for achieving better environmental description capabilities, thereby enhancing the field of view (FOV) of LiDAR sensing. In traditional methods, only real-time 3D point clouds are used to further detect NLOS satellites. However, the ability to detect NLOS satellites is limited by the FOV of the 3D LiDAR sensor 110. In this regard, the second embodiment of the present disclosure improves the shortcomings of the existing technology by accumulating real-time point clouds into SWM. This can effectively enhance the FOV of the 3D LiDAR sensor 110.

[0049] References Figure 3Figure 1 shows the perceived digital world, which shows the difference between the real-time 3D LiDAR point cloud and the SWM. The perceived environment opens a new window to reveal the effects of occlusion or reflection of GNSS signal transmission caused by surrounding high-rise buildings, etc. The 3D LiDAR sensor 110 has a limited FOV, and a single-frame 3D point cloud can only effectively detect low-lying parts of buildings and vehicles (such as double-decker buses). The visibility of high-elevation satellites cannot be effectively classified simply based on the real-time 3D point cloud. The real-time 3D point cloud is also sparse due to the physical scanning angle distribution of the 3D LiDAR sensor 110.

[0050] The above problems can be effectively improved by using the SWM disclosed in the present invention. Figure 3 As shown in (removing ground points from SWM for effective NLOS detection). With the help of SWM, the cut-off elevation angle can be achieved to 76°, so the visibility of satellites with elevation angles less than 76° can be classified. The point cloud in SWM is much denser than the original real-time 3D point cloud, which can significantly improve the accuracy of NLOS detection. A snapshot of the complete SWM map is shown in Figure 3 Buildings and dynamic objects such as double-decker buses and even trees are included in the SWM that were not included in the 3D building model.

[0051] SLAM methods are commonly used to generate point cloud maps based on real-time 3D point clouds. While satisfactory accuracy can be achieved with low drift in a short period of time, errors accumulate over time, leading to significant errors after long-term travel and often preventing loop closure. Therefore, in practical applications, only objects within a circle with a radius of approximately 250 meters are considered to trigger GNSS NLOS reception, while distant buildings are ignored.

[0052] Reference Figure 4 , shows the transformation of the coordinate system. The ECEF coordinate system is fixed to the center of the earth. The first point is selected as the reference of the ENU coordinate system, while the external parameters between the LiDAR, AHRS and GNSS receivers are fixed and calibrated in advance. The L coordinate system (ENU) is obtained by Convert to G coordinate system (ECEF).

[0053] This disclosure uses only the last Nsw coordinate system of the 3D point cloud to generate the sliding window map. As a result, the drift error of the map generation is limited to a small value and is determined by the size of the window (Nsw). In order to minimize the obvious drift in the vertical direction in urban canyons with a large number of dynamic objects, this disclosure uses the absolute ground to constrain the vertical drift. The details of the SWM generation algorithm are demonstrated below:

[0054] Input: from epoch tNsw A series of point clouds from +1 to epoch t Acts as an extrinsic parameter between 3D LiDAR, AHRS, and GNSS receivers.

[0055] Output: SWM

[0056] Step 1: Initialization ←Empty

[0057] Step 2: SWM generation:

[0058] First, all point clouds Align to is the local map of the first coordinate system Secondly, use the following equation to convert the local map A point within Transformed to the receiver carrier coordinate system

[0059] Third, the local map is transformed into A point within Converted to ENU coordinate system

[0060] Now return to reference Figure 2 , the second part of GNSS NLOS detection and correction S200B for GNSS pseudorange measurements is set up with four stages. First, GNSS NLOS detection 251 is used for model verification based on satellite visibility classification using SWM, which can effectively identify LOS satellites and NLOS satellites. Next, if a satellite is classified as NLOS, GNSS NLOS correction 252 is performed as a model calibration stage, which re-estimates the GNSS measurements by correcting the NLOS pseudorange measurements (CNLOS). However, if a satellite is classified as NLOS and its reflection point is not within the SWM (FNLOS), it means that the NLOS correction is not available. The third stage is to perform NLOS reconstruction 253, which is a model repair stage by de-weighting the NLOS measurements for further positioning. The GNSS positioning is estimated based on the pseudorange measurements by a least squares algorithm 254.

[0061] GNSS NLOS Detection via Model Validation 251

[0062] No object detection algorithm is required to recover the actual height of the detected dynamic objects or building surfaces. Since SWM only provides unorganized discrete points, a fast search method based on real-time SWM is used to directly detect NLOS reception without the need for object detection process.

[0063] Inputs to the quick search method include: SWM Elevation angle of satellite s Azimuth at epoch t Maximum search distance D thres and a constant increment value Δd pix The fast search method is used to determine the satellite visibility of satellite s. In step 1, in the ENU coordinate system, The 3D LiDAR center is represented by the initial search point. The search direction connecting the GNSS receiver 130 and the satellite is determined by the elevation angle and azimuth angle of the satellite. The SWM is converted into a kdTree structure. Used to find adjacent points. kdTree is a special structure for point cloud processing that can be performed efficiently when searching for adjacent points. In step 2, given a fixed increment value Δd pix , using the following equation, according to Figure 5 The search direction shown in the example moves the search point to the next point

[0064]

[0065] k represents the index of the search point. Calculate the neighboring points (N k ) According to step 3, if N k Exceeds the predetermined threshold N thres , then at the search point There are some map points from buildings or dynamic objects nearby, and the line of sight between the GNSS receiver 130 and the satellite is considered to be blocked. Therefore, the satellite s is classified as an NLOS satellite. Otherwise, repeat steps 2 and 3 above. If kΔd pix >D thres , the direction between the GNSS receiver 130 and the satellite is along the LOS. thres Can be set to define the distance within which positions are considered for NLOS detection. Using a fast search method, satellite visibility can be classified.

[0066] A demonstration of satellite visibility classification results with NLOS and LOS satellites is shown in Figure 6In the figure, the gray circles represent contaminated (multipath effect) GNSS satellites and the white circles represent healthy GNSS satellites. The numbers provided next to each circle represent the elevation angle of each satellite. From this example, an elevation angle of 54° was detected. As mentioned earlier, the maximum cut-off elevation angle is 76°, and the elevation angle based on SWM may be significantly related to the street width. The narrower the street, the higher the cut-off elevation angle can be achieved. However, NLOS satellites with low elevation angles may cause most of the GNSS positioning errors. Even for challenging urban canyons, the present disclosure is provided to detect and recover signal blockages and reflections to achieve highly accurate GNSS positioning.

[0067] GNSS NLOS corrections via model calibration 252

[0068] NLOS correction is performed through SWM-based model calibration. In order to effectively estimate potential NLOS errors, the distance between the GNSS receiver, satellite elevation and azimuth angles need to be based on the NLOS error model. In particular, model calibration includes detecting reflection points corresponding to NLOS reception. Traditionally, ray tracing technology is used to simulate the NLOS signal transmission path to find NLOS reflectors. However, this technology has the disadvantage of high computational power. The present disclosure provides a method that does not produce continuous building surfaces and clear building boundaries. Instead, SWM only provides a large number of dense, discrete, unorganized point clouds. The search for reflectors from SWM is performed using an efficient kdTree structure based on a reflector detection algorithm.

[0069] The inputs to the reflector detection algorithm used to perform model calibration include: SWM Elevation angle of satellite s Azimuth at epoch t and the azimuth resolution α res The output is the nearest reflection point It is the most likely reflecting surface for NLOS satellites.

[0070] In step 1, the search point is initialized at the center of the 3D LiDAR. The search direction is based on the satellite elevation angle and azimuth Sure.

[0071] For a typical signal reflection path, there are two segments. The first segment is the signal transmission from the satellite to the reflector. The second segment is the signal transmission from the reflector to the GNSS receiver 130. Since the reflected signal should have the same elevation angle as the expected directional signal, step 2 of the reflector detection algorithm includes traversing all azimuths from 0° to 360° with an azimuth resolution of α res , the elevation angle is Thus all possible NLOS transmission paths are found.

[0072] In step 3, if the sight distance associated with the direction is point p j If it is blocked, then the point p j may be a reflection point. In particular, if the connection point p j The line of sight to the satellite is not blocked, then point p j is considered a possible reflector and saved to

[0073] In step 4, repeat steps 2 and 3 until α s >360°, thereby identifying all possible reflectors based on the assumption of the same elevation angle.

[0074] Finally, step 5 is based on the shortest distance assumption, from A unique reflector is detected at the shortest distance between the GNSS receiver 130 and the reflector.

[0075] Advantageously, the reflector detection algorithm does not rely on the detection accuracy of building surfaces. The short range assumption effectively prevents overcorrection because only the closest reflector is identified as the only reflector. The potential NLOS delay of satellite s can be calculated as:

[0076]

[0077] The operator ||*|| is used to calculate the norm of a given vector. sec(*) represents the secant function. Due to the sparsity of SWM, while still denser than a 3D real-time point cloud, there are still some satellites for which reflectors cannot be found using SWM. Therefore, if a satellite is classified as NLOS but no reflectors are found within SWM, NLOS reconstruction is recommended.

[0078] NLOS reconstruction for model repair 253

[0079] Satellites with lower elevation angles and smaller SNRs have a higher probability of being contaminated by NLOS errors. Traditional pseudorange uncertainty modeling methods based on satellite elevation angle and signal-to-noise ratio can produce satisfactory performance in open areas, but not in deep urban canyons. In particular, weighting schemes treat LOS and NLOS in the same way, which is undesirable when NLOS has been detected. The present disclosure corrects GNSS NLOS reception by performing NLOS reconstruction, which essentially models the uncertainty of LOS and NLOS using a weighting scheme, in which a scaling factor is added to de-weight the NLOS measurements. The weighting scheme includes: (1) if the satellite is classified as a LOS measurement, a scaling factor is calculated based on the satellite SNR and elevation angle; (2) if the satellite is classified as a NLOS measurement and the pseudorange error is corrected, a scaling factor is calculated based on the satellite SNR and elevation angle; (3) if the satellite is classified as a NLOS measurement but no reflection point is detected, a scaling factor is calculated based on the satellite SNR and elevation angle and the scaling factor K w Calculate the scale factor.

[0080] GNSS Positioning via Least Squares Algorithm 254

[0081] Pseudorange measurements from a GNSS receiver Expressed as:

[0082]

[0083] in:

[0084] is the geometric range between the satellite and the GNSS receiver;

[0085] is the ionospheric delay distance;

[0086] is the tropospheric delay distance; and

[0087] It is the noise caused by multipath effect, NLOS reception, receiver noise and antenna phase correlation noise.

[0088] At the same time, the traditional Saastamoinen model and Klobuchar model are used to compensate for the atmospheric effect ( and ). The observation model of the GNSS pseudorange measurements from a given satellite s is expressed as:

[0089]

[0090]

[0091] in, is with Correlated noise.

[0092] Assume that the GNSS position is obtained from Subtract the NLOS error from Observation function The Jacobian matrix It can be expressed as:

[0093]

[0094] Where m represents the total number of satellites at epoch t.

[0095] The position of a GNSS receiver can be estimated iteratively by weighted least squares as follows:

[0096]

[0097] Among them, W t The weight matrix based on the weights estimated in NLOS reconstruction 253 is represented as follows:

[0098]

[0099] in,

[0100]

[0101] The weights defined to calculate the LOS measurement based on the satellite SNR and elevation angle are as follows:

[0102]

[0103] Wherein, T represents the signal-to-noise ratio threshold; a, A, and F are predetermined.

[0104] Figure 7 Positioning performance before and after application of the 3D LiDAR-assisted GNSS real-time kinematic differential positioning method developed for the first embodiment of the present invention. Traditional GNSS solutions experience an error of approximately 30 meters, leading to lane misjudgment for vehicles traveling in urban canyons. This improved technology can provide robust and accurate positioning results to support the operation of Level 4 autonomous vehicles.

[0105] Experimental results

[0106] To validate the effectiveness of the proposed method, two experiments were conducted in a typical urban canyon with static buildings, trees, and dynamic objects (e.g., a double-decker bus). The first experiment was conducted in Urban Canyon 1, where the street width was 22 meters and the building height was 35 meters. The second experiment was conducted in Urban Canyon 2, where the street width was 12.1 meters and the building height was 65 meters.

[0107] The vehicle is equipped with a u-blox M8T GNSS receiver, which collects raw GNSS measurements at 1 Hz. The 3D LiDAR sensor is a Velodyne 32, configured to collect raw 3D point clouds at 10 Hz. An Xsens Ti-10 INS collects data at 100 Hz. In addition, a NovAtel SPAN-CPT, a combined GNSS (GPS, GLONASS, and BeiDou) RTK / INS (fiber optic gyroscope, FOG) navigation system, provides ground-truth positioning. The FOG has an operational gyro bias stability of 1 degree per hour and a random walk of 0.067 degrees per hour. The baseline between the rover and the GNSS base station is approximately 7 kilometers. All data is collected and synchronized using ROS. The following five configurations were evaluated: (1) u-blox GNSS receiver only; (2) weighted least squares (WLS); (3) weighted least squares excluding all NLOS satellites (WLS-NE); (4) weighted least squares with re-weighting of all NLOS satellites (R-WLS); and (5) weighted least squares with NLOS corrections if a reflector is detected and re-weighting of NLOS satellites if no reflector is detected (CR-WLS).

[0108]

[0109] Table 1: Positioning performance of GNSS SPP in urban canyon 1

[0110] The results of the GNSS positioning experiments using the five methods are shown in Table 1 above. The positioning results using the u-blox receiver have an average error of 31.02 meters and a standard deviation of 37.69 meters. Due to severe multipath and NLOS reception from surrounding buildings, the maximum error reaches 177.59 meters. For WLS, the positioning error is reduced to 9.57 meters with a standard deviation of 7.32 meters. The maximum error is also reduced to less than 50 meters. After excluding all detected non-line-of-sight satellites, the positioning error increases to 11.63 meters, which is worse than WLS. This happens because excessive NLOS exclusion can significantly distort the perceived geometric distribution of satellites. The availability drops slightly from 100% to 96.01%. Therefore

[0111]

[0112] Table 2: Positioning performance of GNSS SPP in Urban Canyon 2

[0113] In the more challenging situation of Urban Canyon 2, it was discovered that some NLOS satellites reflected by buildings taller than 40 meters may not be detected. Using the u-blox receiver, a positioning error of 30.68 meters was achieved, with a maximum error of 92.32 meters. Using WLS, based on raw pseudorange measurements from the u-blox receiver, a GNSS positioning error of 23.79 meters was achieved. Compared to GNSS positioning using data directly from the u-blox receiver, the maximum error increased slightly to 104.83 meters. Excluding all detected NLOS satellites from the GNSS positioning (WLS-NE), the mean and standard deviation increased to 25.14 meters and 23.73 meters, respectively. Due to the lack of satellites for GNSS positioning, the availability of GNSS positioning data dropped to 95.52%, demonstrating again that completely eliminating NLOS in urban canyons is not advisable. With the help of NLOS reconstruction, the 2D error using R-WLS was reduced to 19.61 meters, ensuring 100% availability. Using the CR-WLS method, the GNSS positioning error was further reduced to 17.09 meters. The improvement in the results demonstrates the effectiveness of the proposed 3D LiDAR-aided GNSS positioning method. The maximum error still reaches 71.28 meters because not all NLOS satellites can be detected and mitigated.

[0114] Figure 8 8 is a system diagram of one possible implementation and combination of a 3D LiDAR-assisted GNSS NLOS mitigation method in a vehicle according to an exemplary embodiment of the present disclosure. The vehicle uses LiDAR-assisted GNSS to support the vehicle's positioning using a satellite positioning system. In certain embodiments, the system includes a 3D LiDAR sensor 110, an AHRS 160, a GNSS receiver 130, a processor 810, a memory 820, a user interface 830, autonomous control 840, and a communication interface 850. In certain embodiments, the 3D LiDAR sensor 110, the AHRS 160, and / or the GNSS receiver 130 can be integrated into the processor 810 without departing from the scope and spirit of the present disclosure. In this case, the processor 810 is a dedicated device with built-in components capable of receiving GNSS satellite signals, directions, and / or a local 3D point cloud map.

[0115] The processor 810 can be one or more general-purpose processors, special-purpose processors, digital signal processing chips, application-specific integrated circuits (ASICs), or other processing structures that can be configured to perform one or more of the following methods described above. The processor 810 is communicatively connected to the 3D LiDAR sensor 110, the AHRS 160, and the GNSS receiver 130 for receiving local maps, directions, and GNSS measurements, respectively. The processor 810 can include a memory 820, or communicate with the memory 820 (as a discrete component), for storing positioning data and / or other received signals and retrieving previous keyframe point cloud data for SWM that is useful for enhancing the FOV of the 3D LiDAR sensor 110.

[0116] In certain embodiments, memory 820 may include, but is not limited to, solid-state storage devices, such as random access memory (RAM) and / or read-only memory (ROM), which may be programmable, flash-updatable, and / or the like. Such storage devices may be configured to implement any suitable data storage, including, but not limited to, various file systems, database structures, and the like. Memory 820 may also include software instructions configured to cause processor 810 to perform one or more functions according to the methods of the present disclosure. Thus, the functions and methods described above may be implemented as computer code and / or instructions executable by processor 810. In one aspect, such code and / or instructions may be used to configure and / or adapt a computer (or other computer device) to perform one or more operations according to the described methods. Memory 820 may therefore include non-transitory machine-readable media having instructions and / or computer code embedded therein. Common forms of computer-readable media include, for example, hard disks, magnetic or optical media, RAM, PROM, EPROM, FLASH-EPROM, USB memory sticks, any other memory chip or cartridge, carrier waves, or any other medium from which a computer can read instructions and / or code.

[0117] User interface 830 allows for interaction with the driver, passengers, and / or other individuals controlling the vehicle and may include one or more input devices selected from the group consisting of a touch screen, a touchpad, buttons, switches, a microphone, etc. The driver or person controlling the vehicle may use user interface 830 to activate or deactivate autonomous control 840 or select different operating modes and / or degrees of autonomous driving.

[0118] Autonomous control 840 is an output device for controlling the vehicle based on GNSS positioning. Processor 810 can thus determine the location of the vehicle, static structures, and dynamic objects, enabling the vehicle to adjust its navigation route, speed, acceleration, onboard alarm system, and / or other functions accordingly. Autonomous control 840 can thus implement various intelligent transportation system functions, such as autonomous driving, semi-autonomous driving, and navigation.

[0119] The communication interface 850 provides an interface that allows data to be communicated with a network, a vehicle, a positioning server, a server, a wireless access point, other computer systems, and / or any other electronic device described herein. In certain embodiments, the communication interface 850 may include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication device and / or a chipset (e.g., a Bluetooth device, an IEEE 802.11 device, an IEEE 802.15.4 device, a WIFI device, a WiMAX device, a cellular communication facility, etc.).

[0120] This system diagram is intended to provide a general description of 3D LiDAR-assisted GNSS, but other components and system blocks may be included as needed. One or more components within the system may be integrated or partitioned and may be located in different physical locations within the vehicle or otherwise remotely located in a server or on a network system. Override controls (not shown) or other safety mechanisms may also be provided to allow the driver or passengers of the vehicle to take control of the vehicle in an emergency.

[0121] As demonstrated, the method of the present disclosure provides an effective way to exclude potential GNSS NLOS receptions, with the help of an environment description generated by a 3D LiDAR sensor and LIO, or SWM using a 3D LiDAR sensor and AHRS. The method can achieve accurate positioning even in urban canyons for autonomous driving. The present invention solves the problem that insufficient positioning accuracy hinders the deployment of autonomous driving applications. On top of all these existing positioning solutions, GNSS-RTK is an indispensable method that can provide global reference positioning in sparse scenes. However, due to signal obstruction and reflection leading to GNSS NLOS reception, its accuracy cannot be guaranteed in urban canyons. The present disclosure solves this problem by leveraging the perception capabilities of 3D LiDAR to detect and eliminate potential GNSS NLOS receptions. In certain embodiments, the method provided herein can achieve higher accuracy even in urban canyon scenarios to meet the navigation requirements of autonomous driving.

[0122] The above method can be directly applied to the autonomous driving industry. Specifically, the present invention can be used to provide accurate positioning solutions even in urban canyons. In addition, the method can also be applied to other autonomous systems with navigation needs, such as mobile robots and drones. At the same time, the method can also be used for land mapping in urban canyons. This illustrates a 3D LiDAR-assisted global navigation satellite system with improved positioning performance according to the present disclosure. Obviously, the variants and other features and functions disclosed above or their alternatives can be combined into many other different configurations, devices, apparatuses and systems. Therefore, the present embodiment is to be considered in all respects to be illustrative and not restrictive. The scope of the disclosure is indicated by the appended claims, rather than by the foregoing description, and is intended to include all changes that fall within the equivalent meaning and range of the claims.

[0123] List of abbreviations

[0124] 3D

[0125] 3DMA 3D map construction assistance

[0126] ADAS Advanced Driver Assistance System

[0127] ADV autonomous driving vehicle

[0128] AHRS attitude and heading reference system

[0129] DOP Dilution of Precision

[0130] ECEF Earth-centered Earth-fixed coordinate system

[0131] ENU Northeast Sky

[0132] FOV field of view

[0133] GNSS Global Navigation Satellite System

[0134] GNSS-RTK Global Navigation Satellite System Real-time Dynamic Differential Positioning

[0135] GPS Global Positioning System

[0136] Kdtree k-dimensional tree

[0137] LAMBDA least squares reduction correlation adjustment

[0138] LiDAR Light Detection and Ranging (LiDAR)

[0139] LIO LiDAR-Inertial Odometry

[0140] IMU Inertial Measurement Unit

[0141] LOS line of sight

[0142] NLOS Non-Line-of-Sight

[0143] PCM point cloud map

[0144] ROS Robot Operating System

[0145] RTK real-time dynamic differential positioning

[0146] SLAM simultaneous positioning and mapping

[0147] SWM Sliding Window Map

[0148] t GNSS epoch

[0149] G ECEF coordinate system

[0150] LENU coordinate system

[0151] s Satellite Index

[0152] GNSS receiver

[0153] BI AHRS carrier coordinate system

[0154] BL LiDAR carrier coordinate system

[0155] BR GNSS receiver carrier coordinate system

[0156] ρ pseudorange

[0157] Pseudorange of the satellite at epoch t

[0158] The position of the satellite at epoch t

[0159] Position of the satellite GNSS receiver at epoch t

[0160] δ r，t GNSS receiver clock error

[0161] Satellite clock error

[0162] Signal-to-noise ratio (SNR)

[0163] K w Scaling factor used for weighting

[0164] Satellite elevation angle

[0165] Satellite azimuth

[0166] k Search point index

[0167] N k Number of adjacent points

Claims

1. A method for use by a vehicle for supporting positioning of the vehicle using a satellite positioning system, the method comprising: Receive LiDAR factors and Inertial Measurement Unit (IMU) factors from the 3D LiDAR sensor and LiDAR Inertial Odometry (LIO); Integrate LiDAR and IMU factors using local factor graph optimization to estimate the relative motion between two epochs. Generate 3D point cloud map (PCM) as auxiliary landmark satellite, which is used to provide low elevation angle auxiliary landmark satellite; receiving Global Navigation Satellite System (GNSS) measurements from satellites via a GNSS receiver; detecting GNSS non-line-of-sight (NLOS) reception from the GNSS measurements using the 3D PCM; as well as The GNSS NLOS reception is excluded from the GNSS measurements to obtain surviving GNSS satellite measurements, thereby improving the quality of GNSS measurements used for positioning of an autonomous vehicle.

2. The method of claim 1, further comprising: performing a global navigation satellite system real-time kinematic (GNSS-RTK) float estimation on the surviving GNSS satellite measurements to obtain a float solution; performing ambiguity resolution to obtain fixed ambiguity solutions; as well as A fixed GNSS-RTK positioning solution is determined from the float solution and the fixed ambiguity solution.

3. The method according to claim 2, wherein: The ambiguity resolution is performed by applying the LAMBDA algorithm.

4. The method of claim 2, further comprising: Feeding back the fixed GNSS-RTK positioning solution to the 3D LiDAR sensor and the LIO; as well as PCM correction is performed using the fixed GNSS-RTK positioning solution to correct the drift of the 3D point cloud.

5. The method of claim 2, further comprising: An initial guess for the float solution is obtained using a least squares algorithm on the GNSS measurements.

Images