Method for making high-precision map based on virtual information track model

Abstract

The embodiment of the invention provides a high-precision map making method based on a virtual information track model. The method comprises the steps that road network information is extracted from original map data or surveying and mapping data, and a reference path is generated according to the road network information; a plurality of Frenet local coordinate systems are established with the reference path as the benchmark, and each local coordinate system at least comprises an arc length axis along the reference path, a transverse offset axis perpendicular to the reference path and an elevation axis perpendicular to the road surface; dividing the reference path into virtual information track segments, wherein each local coordinate system comprises a plurality of virtual information track segments; generating a globally unique identification code for each virtual information track segment, wherein the globally unique identification code comprises encrypting at least one part of geographic information of each virtual information track segment; and binding a static attribute parameter associated with the arc length axis for each virtual information track segment. The map manufactured by the method can meet the requirement of automatic driving above L3 level, and meanwhile, meets the requirement for confidentiality of geographic information.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and more specifically, to a method for creating high-precision maps based on virtual information track models. Background Technology

[0002] With the rapid development of autonomous driving technology, high-level autonomous driving (Level 3 and above) is increasingly reliant on high-precision maps. According to the national standard "Classification of Driving Automation for Automobiles" (GB / T 40429-2021), Level 3 (conditional automated driving), Level 4 (highly automated driving), and Level 5 (fully automated driving) systems require real-time acquisition of road geometry and physical parameters with centimeter-level accuracy to support path planning, vehicle control, and dynamic decision-making. Currently, intelligent connected vehicle base maps can be categorized into "road-level," "lane-level," and "driving control-level" based on their accuracy. "Lane-level" maps carry high-precision road surface information, meeting the general needs of Level 3 / Level 4 autonomous driving, with relative positioning accuracy of up to 10 centimeters for facility targets. "Driving control-level" maps further integrate vehicle-to-everything (V2X) dynamic data (such as roadside unit convergence information and vehicle collaboration data) to form real-time driving information on "digital twin lanes," addressing collaborative control issues in Level 5 beyond-line-of-sight machine driving and dense vehicle environments.

[0003] However, due to the strict legal requirements for the confidentiality of surveying and mapping geographic information in my country, publicly available navigation geographic information must undergo nonlinear transformation of the real geographic coordinates of high-precision maps using nationally recognized geographic information confidentiality processing technology before public use. This process hides the true spatial location, and a terminal plugin integrated into the vehicle must match the vehicle's real-time location with the confidentialized map. This solution involves data processing and plugin integration, has a long implementation cycle, and its accuracy only meets the needs of Level 3 autonomous driving. Therefore, the contradiction between the production and use of high-precision maps and the demands of autonomous driving technology development is prominent. Currently, there is no high-precision map solution on the market that meets both the functional requirements of Level 3 and above autonomous driving and my country's geographic information confidentiality requirements. Therefore, a high-precision map production method that meets both the requirements of Level 3 and above autonomous driving and my country's geographic information confidentiality requirements is needed. Summary of the Invention

[0004] Given the shortcomings of existing technologies, this disclosure aims to provide a method for creating high-precision maps based on virtual information track models.

[0005] This disclosure provides a method for creating a high-precision map based on a virtual information track model. The method includes extracting road network information from original map data or surveying data, and generating a reference path based on the road network information; establishing multiple Frenet local coordinate systems based on the reference path, each Frenet local coordinate system including at least an arc length axis along the reference path, a lateral offset axis perpendicular to the reference path, and an elevation axis perpendicular to the road surface; dividing the reference path into virtual information track segments along the arc length axis, each Frenet local coordinate system including multiple virtual information track segments; generating a globally unique identifier for each virtual information track segment, the identifier including encryption of at least a portion of the geographic information of each virtual information track segment; and binding static attribute parameters associated with the arc length axis to each virtual information track segment.

[0006] The high-precision map production method according to embodiments of this disclosure further includes online verification of a vehicle to determine the virtual information track segment in which the vehicle is located; centimeter-level positioning in the Frenet coordinate system based on information detected by multiple sensors of the vehicle; calibration of positioning deviation using roadside information codes and / or handshake areas; and receiving road information from the vehicle and updating the dynamic attribute parameters of the virtual information track segment in which the vehicle is located based on the road information, wherein the dynamic attribute parameters are associated with the arc length axis.

[0007] Optionally, in the high-precision map production method according to embodiments of the present disclosure, the information detected by the plurality of sensors includes the vehicle heading angle, which is compared with the tangential direction of the arc length axis to determine the deviation of the vehicle in the direction of travel.

[0008] Optionally, in the high-precision map production method according to embodiments of this disclosure, encrypting at least a portion of the road geographic information for each virtual information track segment includes irreversibly encrypting the at least a portion of the geographic information. For example, the encryption process can be implemented using the SM3 algorithm.

[0009] Optionally, in the high-precision map production method according to the embodiments of this disclosure, the plurality of Frenet local coordinate systems and the globally unique identifier code of each virtual information track segment do not include real geographic coordinates, and the globally unique identifier code of each virtual information track segment corresponds to the position of each virtual information track segment in the Frenet local coordinate system.

[0010] Optionally, in the high-precision map production method according to the embodiments of this disclosure, the static attribute parameters bound to each virtual information track segment include one or more of the following: curvature, slope, lane width and / or lane width range, road facilities, traffic sign information, and driving restriction information corresponding to the mileage position of the arc length axis.

[0011] Optionally, in the high-precision map production method according to embodiments of this disclosure, the dynamic attribute parameters associated with the arc length axis include one or more of the following: road surface friction coefficient, traffic flow, visibility, road events, and weather conditions.

[0012] Optionally, in the high-precision map production method according to the embodiments of this disclosure, each handshake area corresponds to one of the plurality of Frenet local coordinate systems, which has encrypted arc length axis reference points. When the vehicle travels to the handshake area, the currently accumulated deviation data of the arc length axis and the lateral offset axis are uploaded online, and the Frenet local coordinate system where the vehicle is currently located is reset according to the received calibration parameters.

[0013] The reconstructed coordinate system of the handshake area requires strict encryption. Preferably, the coordinates of the arc-major axis reference point can be encrypted using the SM4 algorithm. The geographical location of the arc-major axis reference point can be stored in the globally unique identifier code of the virtual information track segment where the reference point is located. Because all subsequent node segment roads (including all road nodes between the handshake area and the handshake area) are determined relative to the initial origin coordinate system, the principles of handshake area reconstruction standards and handshake area setting principles are very important.

[0014] Optionally, in the high-precision map production method according to embodiments of this disclosure, the roadside information code is at least set on roads where the navigation satellite system positioning signal is lost or on roads with high curvature. The vehicle can correct its deviation on the arc length axis and lateral offset axis in real time by reading the roadside information code. Roads with high curvature include, for example, overpasses, roundabouts, continuous curves, and mountain roads. The navigation satellite system can be the BeiDou Navigation Satellite System, the Global Navigation Satellite System (GNSS), etc.

[0015] Optionally, according to the high-precision map production method of this disclosure, dividing the reference path into virtual information track segments along the arc length axis includes dividing the virtual information track segments according to the acquisition nodes of the original map data.

[0016] Compared with the prior art, the embodiments of this disclosure have the following beneficial effects:

[0017] First, the map data is highly confidential. The entire process of creating and applying high-precision maps involves absolutely no storage of real geographical locations. Core data such as Virtual Information Tracks (VITs) are encrypted. Because the Frenet coordinate system is a locally parametric coordinate system based on a "reference path," its axes have no real directional orientation. Furthermore, the segmented encoding of the virtual information track is the "core index" of the map data, designed according to the principles of no real geographical coordinates and "irreversible encryption." Therefore, even if the map data is leaked, it is impossible to associate the encoding with specific roads, meaning it is impossible to obtain the real geographical location through the encoding.

[0018] Second, the map is highly compatible with autonomous driving. The Frenet coordinate system is naturally suited to the "motion control logic" of autonomous driving, which requires only "local relative parameters" without the need for real coordinates. The vehicle achieves precise positioning within the Frenet coordinate system through multi-sensor fusion between multiple sensors (e.g., at least two of inertial measurement units, vision cameras, and LiDAR), without relying on real coordinates. All parameters of the virtual information track segment (e.g., curvature, slope, lane width, dynamic friction coefficient, etc.) are stored in the form of "attributes bound to the Frenet coordinate system." After fusing multiple types of information, the vehicle can directly read the parameters and transmit them to the control module after matching the current segment through the virtual information track encoding, without any "real coordinate calculation" steps.

[0019] Third, the embodiments disclosed herein can achieve centimeter-level accuracy and are stable and reliable. The roadside information code is mainly used for vehicle repositioning in special areas such as navigation satellite system positioning loss environments, while the handshake area is used to correct the error accumulation problem in long, continuous road sections for autonomous driving. By strictly fitting the road morphology with the Frenet coordinate system, and in conjunction with the calibration mechanism of the roadside information code and the handshake area, the cumulative positioning error of the vehicle across the entire road section can be controlled within a few centimeters, far superior to the ±15cm of traditional solutions. Even in long-distance, complex scenarios, the calculation of map parameters (curvature, slope, etc.) remains accurate, with no cumulative error.

[0020] Fourth, the embodiments of this disclosure can achieve navigation support for various complex road scenarios (intersections, roundabouts, interchanges, etc.) through flexible segmentation methods. Optionally, the segments of the virtual information track can be established based on the segmentation nodes determined by the map manufacturer during static map construction, such as nodes established at one-second intervals during static data collection. Alternatively, segmentation rules can be set according to the complexity of the road environment to adaptively segment the virtual information track, so that shorter segment lengths are used in areas with complex road geometry, while longer segments are used in areas with simple geometry, in order to balance modeling accuracy and data efficiency.

[0021] In these scenarios, autonomous vehicles can directly obtain accurate path guidance and control references from maps, without being limited by the problem of traditional maps losing accuracy in special scenarios.

[0022] Fifth, the virtual information track segmentation coding method adopts a joint coding approach that combines regional road segment identification with static information (such as road curvature and slope) and dynamic information (such as traffic control). The coding design can optionally follow the principle of "fixed fields and algorithm generation," and can be generated directly based entirely on existing collected map data without manual intervention or complex encryption processes. It can be directly embedded into existing coding generation processes. The calibration logic of the virtual information track (handshake area calibration, dynamic calibration, etc.) is entirely based on the "local relative parameters" of the Frenet coordinate system, without involving real coordinate transformation. It can reuse the "geometric accuracy verification" and "parameter adaptation" processes in existing map production, making the process simple and without adding new complexity.

[0023] Implementing any apparatus or method of this disclosure does not necessarily require achieving all of the advantages described above simultaneously. Other features and advantages of this disclosure will be set forth in the following description and will be apparent in part from the description and embodiments, or may be learned by practicing this disclosure. The objects and advantages of embodiments of this disclosure may be realized and obtained by means of the structures pointed out in the description, claims, and drawings. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments will be briefly introduced below. Obviously, the drawings described below only relate to some embodiments of this disclosure, and are not intended to limit this disclosure.

[0025] Figure 1 This is a schematic diagram of a coordinate system for constructing a virtual information track model according to an embodiment of the present disclosure;

[0026] Figure 2 This is a flowchart of a method for constructing a virtual information track model based on a coordinate system and generating an offline map according to an embodiment of the present disclosure;

[0027] Figure 3 This is a flowchart of a method for online map updating according to an embodiment of the present disclosure;

[0028] Figure 4A This is a schematic diagram of the distribution of handshake areas on a physical road according to an embodiment of the present disclosure;

[0029] Figure 4B Is with Figure 4A A corresponding virtual information track map and a schematic diagram showing the distribution of handshake areas on it. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. Various different embodiments can be combined with each other to constitute other embodiments not shown in the following description. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.

[0031] Unless otherwise defined, the technical or scientific terms used herein shall have the ordinary meaning as understood by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” and similar terms used in this disclosure and the claims do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Similarly, the terms “an” or “a” and similar terms do not necessarily indicate a quantity limitation. The terms “comprising” or “including” and similar terms mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. The terms “connected” or “linked” and similar terms are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. The terms “upper,” “lower,” “left,” and “right” are used only to indicate relative positional relationships, and these relative positional relationships may change accordingly when the absolute position of the described object changes.

[0032] Figure 1 This is a schematic diagram of a coordinate system for constructing a virtual information track model according to an embodiment of this disclosure. Figure 1 As shown, the Frenet coordinate system of the virtual information track model is a parametric coordinate system established along the reference path. The virtual information track model is a virtual digital road information model that integrates static and real-time dynamic parameters of the road, using a local coordinate system that does not contain real geographical locations as its framework.

[0033] The coordinate system can be defined with the vehicle's location at the road entrance as the starting point, the arc length along the reference path of the road as the s-axis, the horizontal axis perpendicular to the s-axis pointing to the right as the d-axis, and the elevation axis perpendicular to the road surface as the z-axis (not shown in the figure), thus forming the Frenet geometric space.

[0034] Each segment is treated as a "road segment unit," with information such as lane width, curvature, slope, and pavement mechanical parameters marked with centimeter-level precision. The resulting coordinate system is detached from the actual geographical location and serves only for path tracking and control in autonomous driving.

[0035] Reference paths can be extracted from existing map data. Reference paths preserve relative geometry and may not contain any absolute geographic location information. A preferred reference path is the centerline of a lane, or it can be the centerline of a lane, lane boundary lines, etc.

[0036] The following explains the main parameters of the Frenet coordinate system:

[0037] S-axis (arc length axis): The S-axis is established along the cumulative arc length of the reference path. Optionally, the unit of the S-axis can be meters (m). An origin O (s=0) is set at the starting point of the road, and the arc length increases cumulatively along the direction of vehicle movement. For example, a point on the reference path 35.2 meters from the origin O has an S-coordinate of 35.2m.

[0038] The d-axis (lateral axis): Established perpendicular to the s-axis, the d-axis represents the lateral offset of the lane. Optionally, the unit of d can be centimeters (cm), decimeters (dm), etc. On the reference path, d=0; for example, the right side of the reference path can be defined as the positive direction and the left side as the negative direction. For instance, when a vehicle is 15 centimeters to the right of the lane centerline, its d value is +15cm. The d-axis describes the lane width range and the vehicle's lateral position relative to the reference path.

[0039] The z-axis (elevation axis): a height axis perpendicular to the road surface upwards. Optionally, the unit of the z-axis can be set according to the actual road section requirements, such as centimeters (cm), decimeters (dm), etc. At the starting point of the road, z=0, which describes the vertical height corresponding to the s-axis position and is the integral result of the slope along the s-axis. On sloping road sections, the value of the z-axis can be used to calculate the road surface slope i.

[0040] Based on the above coordinate system, the following parameters can be further obtained:

[0041] The θ-axis (heading angle) refers to the angle of deviation of the vehicle's actual direction of travel relative to the tangent of the s-axis. The unit of the θ-axis is degrees (°). When the vehicle is traveling in the s-axis direction, θ = 0°. Clockwise deflection can be set as a positive angle, and counterclockwise deflection can be set as a negative angle. θ is used to correct the vehicle's heading attitude in real time, and works in conjunction with the positioning of the s-axis and d-axis to improve path tracking accuracy.

[0042] Road curvature refers to the degree of curvature of a road on a horizontal plane, usually expressed as "1 / R", where R is the radius of the road curve. The greater the curvature, the more curved the road, and the larger the steering angle required for the vehicle to travel. Road curvature k(s) is a static description of the curvature of the reference path along the s-axis; it is the rate of change of the s-axis tangent heading angle. Road curvature indirectly provides a calibration target for the heading angle by generating the s-axis tangent heading angle (static reference).

[0043] As mentioned above, the road surface slope (i) is the static relationship between the z-axis (elevation axis) and the s-axis (arc length axis) in the Frenet coordinate system, defined as the "derivative of elevation with respect to the arc length of the s-axis" (i = dz / ds). For example, an uphill road surface slope of i = 3% means that the elevation rises by 3 meters for every 100 meters of forward travel. The static slope can directly determine the distribution pattern of elevation.

[0044] By constructing a virtual information track model based on the Frenet coordinate system, all positioning and road parameters can be closely related to the reference path and directly correspond to the vehicle's driving path. This makes it easier to obtain the vehicle's lateral offset information without the need for complex coordinate transformations or curve calculations on the vehicle side, greatly improving the efficiency of map data utilization in autonomous driving systems.

[0045] Figure 2 A flowchart illustrating a method for constructing a virtual information track model based on a coordinate system and generating an offline map according to an embodiment of the present disclosure is shown.

[0046] In step S101, firstly, map information, including road network data, is collected from static maps or surveying data, and a reference path is generated from the road network data. Specifically, original map information is collected from original static electronic map data conforming to national standards or surveying data obtained from surveying departments, such as CGCS2000 / WGS84 compliant electronic maps. A reference path is generated based on the road network data in the map. As mentioned above, the reference path can preferentially select lane centerlines. Optionally, in this step, the actual geographical location in the original map data or surveying data can be encrypted, for example, using SM4 symmetric encryption, which can hide the actual geographical location in subsequent map production processing.

[0047] In step S102, parametric modeling is performed based on the Frenet coordinate system. The road network data is transformed into the Frenet local coordinate system, a virtual information track is generated in the local coordinate system, and virtual information track segments are established.

[0048] The Frenet coordinate system is a parameterized coordinate system established along a reference path, as explained above. Based on the parameter definition of the Frenet coordinate system, road network data is transformed into multiple continuous Frenet local coordinate systems, thereby constructing a virtual information track. Specifically, using the reference path as a reference, the origin of each local coordinate system is established, i.e., s=0, d=0, z=0. The position of the origin on the road can be determined along the arc length on the s-axis; the specific method for determining the origin will be explained in the description of the handshake area below. The reference point coordinates (origin) of each local coordinate system are encrypted and stored, for example, using an irreversible encryption algorithm, preferably the SM4 algorithm. The core advantage of SM4 encryption for reference point coordinates lies in the fact that SM4 is a symmetric encryption algorithm independently designed in my country (belonging to the national cryptographic algorithm system). The encryption process for reference point coordinates is simple and efficient, with low power consumption. The encrypted data is tamper-proof and anti-eavesdropping, with strong data confidentiality, compliance, and interoperability.

[0049] The encryption processes mentioned in this disclosure can preferably use national cryptographic algorithms (including SM2, SM3, SM4, etc.), and are not limited to the specific examples given in this disclosure. The appropriate algorithm can be selected as needed. For example, a hash algorithm can be used first to encrypt the unique code before encrypting it with a key, or geographic data can be encrypted separately using SM4-GCM and accompanied by an authentication tag. National cryptographic algorithms are commercial cryptography recognized by the State Cryptography Administration, comply with national legal requirements, and can avoid compliance risks (the handshake area reference point coordinates are core data supporting high-level autonomous driving safety; encryption using national cryptographic algorithms is a "legal requirement," avoiding project risks due to non-compliance).

[0050] In step S103, after establishing the virtual information track, each local coordinate system is divided into multiple segments along the s-axis, serving as virtual information track segments. The virtual information track segmentation can adopt the node segmentation rules from the collected original map (e.g., using the number of data collection nodes determined per second during map data collection as the nodes for virtual information track segmentation), or it can be segmented according to fixed arc length intervals (e.g., segmenting with arc lengths of 5 meters or 10 meters), or it can be segmented according to different segmentation principles based on road complexity. When segmenting based on road complexity, the arc length of each segment unit can be different depending on its road characteristics. For example, if the road curvature changes drastically (e.g., sharp bends, roundabouts), the segment length can be shortened as needed to improve fitting accuracy.

[0051] For example, starting from the origin, the first segment covers s∈[0, Δs1], the second segment covers s∈[Δs1, Δs2], and so on, thus dividing the entire road into several continuous arc-length segments, each corresponding to a certain mileage interval on the road. By establishing the aforementioned local coordinate system starting from the road origin and further segmenting it, multiple virtual information track segments are generated.

[0052] Information for each virtual information track segment is encoded. This information may include lane information, road facility information, etc. To distinguish each virtual information track segment, a unified road coding system is used to globally encode each segment, ensuring that each code does not contain any explicit geographical location, but is only associated with the Frenet origin coordinates of the local coordinate system or encrypted data. Specifically, a globally unique identifier code (VIT-ID) is generated for each virtual information track segment. This code uses a multi-field combination design. For example, the VIT-ID may include VIT identifier, administrative division code, road sign, road segment number, lane sign, driving direction, etc. For virtual information track segments corresponding to the handshake area, the VIT-ID includes the encrypted reference point coordinates of the handshake area. The VIT-ID may also include a checksum, which can be used to verify the integrity of the encoded data and prevent tampering. Some or all of this information can be encrypted as necessary through encoding. Since the Frenet local coordinate system replaces the original geographical location, i.e., it does not include the original geographical location, the VIT-ID cannot reconstruct the original geographical location. For example, random salting and multi-round iterative hashing algorithms can be used to encrypt some geographical information. Optionally, important information (such as information related to roads and the S-axis) can be irreversibly mapped using encryption algorithms, or the two pieces of information can be associated and encrypted. Through this key-driven irreversible mapping, third parties cannot deduce the actual geographical location of the road or construct a readable map from the VIT-ID data, thus fully ensuring the confidentiality of the map data. Continuous segments within each road can also be identified using segment indices in the VIT-ID.

[0053] In step S104, static and dynamic attributes related to the s-axis position can be bound to each VIT segment. Static attributes mainly include road physical properties and traffic control information, which can be obtained from the original static map or survey data and converted into information associated with at least the s-axis coordinate during the offline phase. Static attributes can be pre-configured offline and bound to the corresponding mileage position of the s-axis in the local coordinate system as additional attributes. Physical attribute information can include curvature, slope, lane width, etc. For example, the following parameter can be attached to the s-axis position of each segment unit: curvature k(s) (e.g., k=0 for straight road segments, and a curvature function or average value, such as 1 / 50m, for curved road segments). -1The static elements include the slope i (e.g., calculated using i=ΔZ / Δs), lane width range (e.g., the lane d∈[-175, 175]cm for this segment), and traffic control information such as traffic light locations (e.g., a traffic light at s=50m in a segment, with a red light duration of 25 seconds and a green light duration of 35 seconds), speed limits, and turning restrictions (e.g., left turns are prohibited in the s∈[1200,1300]m range). These static elements are stored in association with specific s-axis positions through key-value pairs or structured data, allowing autonomous vehicles to directly search for relevant road information based on their current location. This information can be attached to offline maps as static attributes or updated in real-time as dynamic attributes through real-time collection and calculation (see the description of dynamic road attributes below). The parameters of the static attributes are associated with the s-axis arc length and can be directly accessed by the autonomous driving system.

[0054] Dynamic attributes can also include information such as the physical properties of the road and traffic control. They can also include other driving-related information, such as visibility, traffic flow, and road surface friction coefficient. Information collected in real-time by online vehicles or road management agencies can be sent to a cloud-based control system (e.g., a "virtual information track" network). The control system generates or updates dynamic attributes based on the collected information and sends these attributes to the vehicle based on its location, providing real-time updates to the road's dynamic attributes. The parameters of the dynamic attributes are related to the S-axis arc length and can also be directly accessed by the autonomous driving system.

[0055] Figure 3 A flowchart illustrating a method for online real-time map updating and vehicle positioning according to an embodiment of this disclosure is shown. This method achieves precise S / D axis positioning of vehicles and VITs, cross-road segment calibration, and dynamic information fusion.

[0056] In step S201, the autonomous vehicle performs VIT matching and key verification with the cloud. After powering on or entering a certain area, the autonomous vehicle requests a list of VIT indexes near its current location from the cloud's control system. Based on the starting point coordinates of the VIT segments, the autonomous vehicle uses its positioning to initially filter out one or more possible VIT segments, for example, filtering candidate VIT segments based on navigation satellite system positioning. Subsequently, the vehicle uses a pre-loaded local key to verify the ID of the candidate VIT segment, for example, by sending the local key and VIT-ID to the cloud for authorization verification. After successful verification, the cloud confirms the VIT segment corresponding to the autonomous vehicle based on the starting point coordinates of the initially filtered VIT segments. It returns the coordinate information and / or offline map information of the target VIT segment in the Frenet coordinate system, such as the s-axis range, d-axis lane parameters, and static attributes bound to the segment. None of this data contains any real geographical location. Through this process, the vehicle can securely obtain the required high-precision map information, and only authorized vehicles can interpret the complete VIT content, further improving the security of data use.

[0057] In step S202, the vehicle's positioning in Frenet coordinates is calculated, and the positioning deviation is calibrated using roadside information codes and / or the handshake area. This positioning deviation calibration primarily addresses the s-axis coordinate deviation, while also calibrating other parameters in the coordinate system. After acquiring map information for the corresponding VIT segment, the vehicle utilizes its multi-sensor fusion and pose estimation technologies to calculate its own coordinates relative to the reference path within the Frenet coordinate system, achieving centimeter-level positioning.

[0058] Alternatively, when positioning the S-axis, the relative position of the vehicle in the local coordinate system can be determined by measuring the incremental arc length of the travel using the vehicle's wheel odometer.

[0059] To eliminate accumulated errors, visual sensors can be used to identify pre-deployed roadside information codes along the road (e.g., spaced QR codes, RFID tags, visual signs, ground magnetic markers, etc.), which contain precise current positioning values ​​(e.g., S-axis values). Vehicles use visual cameras (front / side cameras), RFID readers, or magnetic sensors to read the roadside information codes and precisely calibrate their S-axis position, achieving an accuracy of ±1cm. For example, if the odometer calculates an S-axis position of 50.15m, but the roadside information code indicates the true S-axis reference should be 50.00m, the vehicle immediately corrects its S-axis to 50.00m, thus eliminating odometer drift error. The roadside information codes can also carry the current VIT segment identifier for vehicle identification, allowing vehicles to quickly confirm their segment without frequent cloud requests, improving positioning efficiency and reducing communication latency.

[0060] The following is a detailed explanation of the roadside information code.

[0061] Roadside information codes are passive / low-power physical coding devices (which can take the form of QR codes, anti-interference barcodes, RFID tags, laser reflective codes, magnetic signs, etc.) deployed at the edges of roads (such as on the ground, guardrails, and beside curbs). Roadside information codes can be set at fixed intervals, or according to the complexity of the road (e.g., denser settings can be used in areas with large curvature changes such as curves and roundabouts), or according to an "on-demand deployment" scheme (as described below). Roadside information codes only serve positioning calibration in the Frenet coordinate system.

[0062] The roadside information code can include local parameters corresponding to segments of the virtual information track. Its design strictly adheres to the principle of "not disclosing real geographic information and only serving Frenet positioning." All fields are derived from pre-set parameters during the offline construction phase, containing no sensitive information such as geographic location, real road name, or mileage marker. Even if the code is maliciously obtained, it is impossible to deduce the road's true location. All data for the roadside information code is pre-built during the offline construction phase. Vehicles do not need to connect to the internet when reading it; data is collected and analyzed in real time through local sensors. This ensures a positioning latency of less than 50ms while avoiding the risk of data leakage during transmission.

[0063] Roadside information codes can contain core essential information and, as needed, other extended information. The core essential information includes fundamental fields that ensure positioning calibration and matching; these are static data pre-built offline within the code, and can contain only local static parameters of a single segment that cannot be modified. For example, a roadside information code includes VIT segment identifiers and S-axis reference values ​​(with high precision), which are fully aligned with the VIT model and strictly consistent with the parameters of the offline-constructed VIT segments. Furthermore, roadside information codes can also include consecutively numbered coded signs within the same road, with each number corresponding to one roadside information code for that road. Roadside information codes can also incorporate built-in verification information; after a vehicle reads the code, it can locally verify data integrity (if the checksum matches, it confirms the code has not been tampered with), ensuring calibration accuracy.

[0064] The "on-demand deployment" scheme for roadside information codes described above is only implemented in high-risk positioning areas, such as when navigation satellite systems lose positioning or in special scenarios like high-curvature road sections (e.g., overpasses) or continuous tunnel groups. This achieves the following effects:

[0065] (1) Construction costs are significantly reduced: the total number of deployments can be reduced by more than 80% (for example, only 20-50 information codes are needed for 100 kilometers of roads), significantly reducing equipment procurement and maintenance costs;

[0066] (2) More focused on specific scenarios: Prioritizes coverage of high-risk positioning areas such as navigation satellite system positioning loss, visual limitations, and high curvature turning, thereby improving resource utilization efficiency;

[0067] (3) Improved maintenance convenience: The information codes are mainly distributed in sections with clear structure and centralized maintenance, which facilitates periodic inspection and replacement.

[0068] Roadside information codes are offline pre-built "local coordinate calibration anchor points." Their core function is to assist autonomous vehicles in accurately calibrating their S-axis (arc-length axis along the reference path) positioning in the Frenet coordinate system, eliminating accumulated errors from sensors such as wheeled odometers and IMUs, while avoiding the leakage of any real geographical information. Roadside information codes can be mainly used when vehicles are in areas without network coverage (such as tunnels) or when cloud interaction latency is too high (such as during peak-hour communication congestion). They can independently complete S-axis calibration without relying on the cloud, ensuring uninterrupted positioning.

[0069] When locating along the d-axis, sensors such as onboard cameras and LiDAR can be used to identify lane lines or road edges, measure the lateral distance of the vehicle deviating from the reference path, and calculate the d-axis offset with an accuracy of ±2cm. Combined with lane width data for that road segment on the map, the lateral position of the vehicle can be monitored to ensure it remains within a safe range.

[0070] When locating along the θ-axis, a high-precision IMU can be used to measure the vehicle's heading, which is then compared with the s-axis tangent of the reference path on the map. The deviation θ between the vehicle's heading angle and the s-axis is calculated in real time, and the vehicle's attitude is corrected accordingly. Typical accuracy can reach ±0.1°. By combining s, d, and θ positioning, the vehicle can achieve real-time centimeter-level positioning in the Frenet coordinate system.

[0071] In this method, to prevent small errors from accumulating gradually after the vehicle has traveled a long distance, this disclosure also sets up a periodic cloud calibration mechanism – the “handshake zone”. Figure 4A This is a schematic diagram of the distribution of handshake areas on a physical road according to an embodiment of the present disclosure; Figure 4B Is with Figure 4A The corresponding virtual information track map and its handshake zone distribution diagram are shown. The handshake zones are used for coordinate reset and position calibration. As shown in the figure, the handshake zones are spaced along the S-axis, and the coordinates are reset when a vehicle passes through a handshake zone. The intervals of the handshake zones correspond to the local coordinate system.

[0072] The handshake area can be set up according to the following principles:

[0073] (1) Dynamically adapt to road complexity and encrypt the handshake zone configuration for complex scenarios. Specifically, the cumulative error rate varies for different road scenarios, so the interval settings need to be differentiated to ensure that calibration is completed before the error exceeds the threshold. On conventional roads, such as urban sections with low curvature, the cumulative rate of vehicle heading angle drift and d-axis offset is slow. Even with a larger interval, calibration can be completed before the error exceeds the threshold. For example, a handshake zone with an interval of 10km corresponds to a local coordinate system s-axis length of 10km. The handshake zone can be set at the junction of different types of roads, different administrative regions, etc., to simultaneously calibrate the coordinate system deviation across regions. For complex roads, such as continuous curves, overpasses, roundabouts, etc., the error accumulation will be faster, so a handshake zone with a smaller interval is required.

[0074] (2) The handshake zone needs to be set so that there are no calibration blind spots on the entire road section, so as to achieve positioning calibration coverage throughout the entire life cycle of the road and avoid loss of control due to accumulated errors.

[0075] (3) The handshake area and the roadside information code complement each other to form a system that combines "low-frequency global calibration" and "high-frequency local correction". The roadside information code can be used to ensure uninterrupted positioning and is responsible for high-frequency local correction in high-risk positioning areas. The handshake area is responsible for low-frequency global calibration in the interval to solve the problem of large error accumulation. The two work together to cover the calibration needs of the whole scenario.

[0076] (4) The handshake area must have clear physical markings and its material is suitable for long-term stable operation outdoors, so that the vehicle sensor can quickly identify and trigger the reconstruction process, avoiding calibration omissions.

[0077] Based on the above principles, the nodes of the handshake zone correspond to the local coordinate system and are set at intervals along the s-axis. For example, a handshake zone is set at the junction of adjacent local coordinates. The information in the handshake zone does not include the actual coordinates; it only stores the encrypted reference markers for that location in the cloud, and clearly identifiable markers can be deployed on the corresponding roadside. When a vehicle approaches the handshake zone, the onboard system can upload the accumulated positioning error information within the current VIT segment to the cloud, for example, reporting "s-axis deviation Δs = +0.15m, d-axis deviation Δd = -0.08m". The cloud calculates the calibration parameters (e.g., the s, d, and z values ​​should be calibrated to 0) based on the pre-saved encrypted s-reference points of the handshake zone and sends these parameters to the vehicle. When the vehicle passes the handshake zone marker, calibration is triggered. The vehicle resets its own coordinates and adjusts the θ-axis direction to be consistent with the tangent direction of the new road segment reference path. By utilizing the "reset" function of the handshake area, accumulated errors generated during long-distance, segmented driving can be completely eliminated, ensuring positioning accuracy across road types (e.g., the intersection of urban arterial roads and expressways), administrative districts, or complex road junctions (e.g., interchange entrances / exits, roundabout entrances / exits). In this disclosed solution, the combination of high-frequency local calibration (road test information code) and low-frequency global calibration (handshake area) maintains positioning accuracy within ±10cm across the entire road segment. This method combines the advantages of both high-frequency and low-frequency calibration.

[0078] The s-value of the roadside information code is the Frenet arc length, and the encrypted control point of the handshake area is a "fuzzy reference without real coordinates" (which can only be solved in the cloud and does not send the real geographical location to the vehicle), which fully complies with the national geographic information confidentiality requirements and does not have the problem of leaking sensitive information.

[0079] Optionally, to indicate approach to the handshake zone and avoid coordinate jumps during handshake zone calibration, at least one roadside information code can be deployed near the handshake zone (e.g., approximately 100m ahead). Before the vehicle enters the handshake zone, the S-axis deviation can be corrected using the roadside information code (e.g., correcting s=999.85m to 999.98m), and then global calibration can be completed upon entering the handshake zone, further improving reconstruction accuracy.

[0080] In step S203, based on the VIT segmentation and road information obtained from the vehicle / roadside unit (RSU), the cloud updates static and / or dynamic attributes and sends them to the vehicle. When the roadside unit uploads data to the cloud, it can also verify the data using the aforementioned key and fixed code matching method.

[0081] As mentioned above, static attributes can include curvature, slope, lane width, traffic light position, and steering restrictions. Static attributes are relatively stable and generally do not require frequent updates. When the cloud detects a change in a static attribute, it updates the stored static attributes in real time. Dynamic attributes can include road, traffic, and environmental information that changes over time. For example, the RSU collects dynamic information related to the S-axis in real time, fuses it to the VIT segment using navigation satellite systems / IMUs / odometers / vision / radar, and uploads the measurement results. The cloud can fuse, verify, and update the dynamic information. After verifying the validity of the information, the dynamic attributes of the corresponding VIT segment are updated, and the information can be distributed to all vehicles within the segment via the vehicle wireless communication network. The dynamic information received by the cloud can include, for example, roadside unit information, road surface positioning information, navigation satellite system positioning location information, vehicle and pedestrian dynamic information, road scene information, traffic control information, and meteorological information. Valid information can be extracted from these sources, fused, generated, and / or updated to enhance the dynamic attributes of the VIT segment, thereby enabling high-level autonomous driving through information-based decision-making. For example, dynamic information can include the road surface friction coefficient μ (e.g., if a road segment is detected to be slippery due to rain, the μ value within that segment decreases), traffic flow, visibility (e.g., global visibility of 800m), and other event information such as accidents, construction, and congestion. Initial values ​​for dynamic attributes can be set to default values ​​during the offline phase, and are corrected by online data in actual applications. All dynamic attributes are managed using S-axis mileage segments as indexes. During vehicle operation, the latest dynamic parameters can be directly obtained based on the VIT segment to which the vehicle currently belongs, enabling timely responses to environmental changes.

[0082] As can be seen from the above embodiments, the method disclosed herein can effectively construct and apply high-precision maps in real-world road environments. In the offline phase, the original map is encrypted and parameterized to generate a virtual information track that meets confidentiality requirements; in the online phase, dynamic attributes are updated using information collected by the RSU (Roadside Unit). Vehicles utilize this VIT (Virtual Information Technology) to achieve precise positioning and dynamic information acquisition. The vehicle's positioning error can be maintained at the centimeter level throughout the entire journey, and it can obtain timely map-level support information even when encountering complex road conditions such as overpasses, traffic lights, and rainfall. Therefore, the method disclosed herein is applicable to high-level autonomous driving.

[0083] This disclosed Virtual Information Track (VIT) is a novel map representation method for autonomous driving control and path planning. This system replaces the traditional latitude and longitude sandbox or absolute geographic coordinate system with a defined Frenet local coordinate system, achieving a direct abstraction of the geometric and physical attributes of roads. In the VIT model, the map no longer centers on "human-readable" geographic locations (latitude and longitude, road names), but is transformed into a "control-usable geometric tensor space," where each coordinate point carries a set of parameters directly usable by the autonomous driving control system. This not only facilitates the autonomous driving system in directly reading the data required for control but also eliminates the dependence on readable geographic locations at the architectural level, thus possessing inherent confidentiality and anti-leakage characteristics.

[0084] Without compromising vehicle functionality, an irreversible / key-controlled mapping method is used to encrypt road codes and geometric information before publishing the VIT (Virtual Information Technology). This prevents any receiving end (without the key) from obtaining the geographic location or generating maps for human identification. The combination of VIT segmentation and encryption methods differs from conventional encryption or simple cut-and-paste public disclosure and represents a usable data encryption strategy for autonomous driving functions.

[0085] The local coordinate system integrates the local geometry, dynamics, and properties (such as friction, roughness, and slope) of the road onto virtual information track segment nodes, supporting centimeter-level accuracy. This allows the controller to directly read the required information and reduces the overhead of real-time geometry solving.

[0086] The embodiments of the present invention can be modified and altered in various ways without departing from the spirit and scope of the invention. Therefore, it should be understood that the scope of protection of the present invention should not be limited to the exemplary embodiments described above, but should cover the full scope defined by the claims and their equivalents.

Claims

1. A method for producing a high-precision map based on a virtual information track model, characterized by, Comprising extracting road network information from original map data or survey data, and generating a reference path according to the road network information; establishing a plurality of Frenet local coordinate systems with the reference path as the reference, each Frenet local coordinate system including at least an arc length axis along the reference path, a lateral offset axis perpendicular to the reference path, and an elevation axis perpendicular to the road surface; dividing the reference path into virtual information track segments along the arc length axis, each Frenet local coordinate system including a plurality of virtual information track segments; generating a globally unique identification code for each virtual information track segment, the globally unique identification code including at least a part of the geographical information of each virtual information track segment being encrypted; binding a static attribute parameter associated with the arc length axis for each virtual information track segment.

2. The method of claim 1, wherein the virtual information track model is a virtual lane model. Further comprising verifying the vehicle online to determine the virtual information track segment where the vehicle is located; performing centimeter-level positioning in the Frenet coordinate system according to the information detected by a plurality of sensors of the vehicle; calibrating the positioning deviation using a roadside information code and / or a handshake area; and receiving road information from the vehicle and updating the dynamic attribute parameter of the virtual information track segment where the vehicle is located according to the road information, the dynamic attribute parameter being associated with the arc length axis. The information detected by the plurality of sensors includes a vehicle heading angle, which is compared with the tangential direction of the arc length axis to determine the deviation of the vehicle in the direction of travel.

3. The method of claim 2, wherein the virtual information track model is a virtual lane model. The encryption of at least a part of the road geographical information of each virtual information track segment includes irreversible encryption of the at least a part of the geographical information.

4. The method of claim 1, wherein the virtual information track model is a virtual lane model. The plurality of Frenet local coordinate systems and the globally unique identification code of each virtual information track segment do not include real geographical coordinates, and the globally unique identification code of each virtual information track segment corresponds to the position of the each virtual information track segment in the Frenet local coordinate system.

5. The method of claim 1, wherein the virtual information track model is a virtual lane model. The static attribute parameter bound for each virtual information track segment includes one or more of curvature, slope, lane width and / or lane width range, road facilities, traffic sign information, and driving restriction information corresponding to the mileage position of the arc length axis.

6. The method of claim 1, wherein the virtual information track model is a virtual lane model. The dynamic attribute parameter associated with the arc length axis includes one or more of road surface friction coefficient, traffic flow, visibility, road event, and weather condition.

7. The method of claim 2, wherein the virtual information track model is a virtual lane model. Each handshake area corresponds to one of the plurality of Frenet local coordinate systems, has an encrypted arc length axis reference point, transmits the accumulated deviation data of the arc length axis and the lateral offset axis online when the vehicle travels to the handshake area, and resets the Frenet local coordinate system where the vehicle is currently located according to the received calibration parameter.

8. The method of claim 2, wherein the virtual information track model is a virtual lane model. The roadside information code is at least set on a road where the navigation satellite system positioning signal is lost or a road with high curvature, and the vehicle can read the roadside information code to correct the deviation of the vehicle in the arc length axis and the lateral offset axis in real time. 9.The method of claim 2, wherein, ​ 10.The method of manufacturing a high-precision map based on a virtual information track model according to claim 1, wherein, Dividing the reference path along the arc length axis into virtual information track segments includes dividing the virtual information track segments according to acquisition nodes of the original map data.

Images