Multi-sensor fusion positioning method and device for multi-layer roadway scene, electronic equipment and storage medium

Abstract

The invention relates to a multi-sensor fusion positioning method for a multi-layer roadway scene. The method comprises the following steps: acquiring attitude data of an inertial measurement unit, point cloud data of a laser radar and UWB (Ultra Wideband) beacon data in different time periods; importing the point cloud data and the attitude data into an SLAM algorithm to generate a plurality of first global maps in different time periods; combining the plurality of first global maps to generate a second global map; splitting the second global map to generate a plurality of topological sub-maps; and obtaining vehicle positioning according to the UWB beacon data. The invention further discloses a device, electronic equipment and a storage medium suitable for the method. By using the multi-sensor fusion positioning method and device for the multi-layer roadway scene, the electronic equipment and the storage medium provided by the invention, the problems that the labor cost of field surveying and mapping and instrument measurement is relatively high, and the method is not suitable for large-scale underground scenes are solved; and the laser radar cannot be started at any position.

Description

Technical Field

[0001] The present invention relates to the field of positioning systems, and in particular to a multi-sensor fusion positioning method, device, electronic equipment and storage medium for multi-layer tunnel scenes. Background Art

[0002] High-precision positioning technology is not only the key to the safe driving of autonomous vehicles, but also the basis for autonomous vehicle navigation. Through high-precision positioning technology, autonomous vehicles can accurately know their position on the road, thereby safely avoiding obstacles and accurately performing driving tasks. In the face of multi-layer tunnels in mines, multi-sensor fusion positioning technology faces new challenges.

[0003] In the existing technology, when realizing multi-sensor data fusion positioning, on-site mapping and instrument measurement are often used to obtain the global position of the ultra-wideband beacon, that is, the global position of the UWB beacon. This method has a high labor cost and is not suitable for large-scale underground scenes. Laser radar is also used to achieve local precise positioning, but it cannot achieve startup at any position.

[0004] Therefore, the prior art has defects and needs to be improved and developed. Summary of the invention

[0005] The embodiments of the present invention provide a method for solving the problem that in the prior art, when realizing multi-sensor data fusion positioning, on-site surveying and instrument measurement are often performed to obtain the global position of the ultra-wideband beacon, that is, the global position of the UWB beacon. This method has a large labor cost and is not suitable for use in large-scale underground scenarios. There is also a technical problem that laser radar is used to achieve local precise positioning, but it cannot realize the startup of any position.

[0006] In a first aspect of an embodiment of the present invention, a multi-sensor fusion positioning method for a multi-layer tunnel scene is provided, comprising:

[0007] Collect attitude data of the inertial measurement unit located on the vehicle and used to detect the roadway environment, point cloud data of the laser radar, and data of the UWB beacon arranged in the roadway at different time periods;

[0008] Importing the point cloud data and the posture data into a SLAM algorithm to generate a first global map of a plurality of different time periods;

[0009] Merging a plurality of the first global maps to generate a second global map;

[0010] splitting the second global map to generate a plurality of topological sub-maps;

[0011] The vehicle positioning is obtained according to the UWB beacon data.

[0012] Furthermore, the method of importing the point cloud data and the posture data into the SLAM algorithm to generate a first global map of several different time periods includes:

[0013] Constructing several layers of point cloud maps according to the point cloud data;

[0014] Label each layer of the point cloud map according to the roadway hierarchy;

[0015] Output the last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud maps with adjacent layer numbers in sequence;

[0016] Fusing several layers of point cloud maps to generate the first global map;

[0017] The last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud map with adjacent layer numbers are imported into the ICP algorithm to generate the first global map through point-to-surface ICP.

[0018] Furthermore, the method of merging a plurality of the first global maps to generate a second global map includes: fusing map segments of overlapping areas in a plurality of the first global maps by a map merging method to generate the second global map.

[0019] Furthermore, the method for obtaining the precise location of the vehicle according to the UWB beacon data includes:

[0020] Obtaining a rough position of the vehicle according to the UWB beacon data;

[0021] The rough position of the vehicle, the real-time lidar data of the vehicle, and the second global map are matched to obtain the vehicle positioning.

[0022] Furthermore, the method for obtaining vehicle positioning according to the UWB beacon data further includes: updating the positioning of the vehicle in the topological sub-map according to the posture data of the inertial measurement unit.

[0023] Further, the method for obtaining the rough position of the vehicle according to the UWB beacon data includes:

[0024] According to the reflection intensity of the reflective sticker posted on the UWB beacon, a point P with a reflection intensity value greater than A is obtained, wherein the reflective sticker is provided with label numbers according to the order in which the vehicles enter the lane; wherein the value of A is a critical value for distinguishing the reflective sticker from other reflective objects in the lane environment;

[0025] According to the light intensity point P, a nearest neighbor search is performed using KD-Tree to obtain n points closest to the point P, and the points whose distance from the point P is less than r are clustered into a set Q; wherein n is a natural number greater than, and r is the value of the longest straight line segment on the reflective surface of the reflective tape;

[0026] If the number of elements in the set Q continues to increase, then any point other than the point P is selected from the set Q as the point K, and the previous step is repeated until the number of elements in the set Q no longer increases;

[0027] If the number of elements in the set Q does not increase any more, the nearest neighbor search is terminated to obtain the point cloud intensity of the reflective tape;

[0028] The point cloud intensity of the reflective tape is used to output the three-dimensional coordinates of the center point of the reflective tape and the label number of the reflective tape to obtain the rough position of the vehicle.

[0029] In a second aspect of an embodiment of the present invention, a multi-sensor fusion positioning device for a multi-layer tunnel scene is provided, comprising:

[0030] An acquisition module, used to collect attitude data of an inertial measurement unit located on the vehicle and used to detect the roadway environment, point cloud data of a laser radar, and data of UWB beacons arranged at equal intervals in the roadway at different time periods;

[0031] A first generating module, used for importing the point cloud data and the posture data into a SLAM algorithm to generate a first global map of a plurality of different time periods;

[0032] A second generating module, used for merging a plurality of the first global maps to generate a second global map;

[0033] A third generating module, used for splitting the second global map to generate a plurality of topological sub-maps;

[0034] The acquisition module is used to obtain the vehicle positioning according to the UWB beacon data.

[0035] According to a third aspect of an embodiment of the present invention, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the multi-sensor fusion positioning method for multi-layer tunnel scenes.

[0036] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the multi-sensor fusion positioning method for a multi-layer tunnel scene is implemented.

[0037] Beneficial effects:

[0038] It can be seen from the above scheme that the present invention provides a multi-sensor fusion positioning method, device, electronic device and storage medium for multi-layer tunnel scenes. Through multi-sensor data fusion, the inertial measurement unit, laser radar, and UWB beacon data are fused to achieve high-precision positioning in the complex environment of multi-layer tunnels; through data collection and map generation in different time periods, and real-time update of vehicle posture data, it adapts to the changes in tunnel structure in dynamic environments such as mines, enhances the flexibility of the system, and ensures the dynamic response capability and accuracy of the positioning system; through systematic scheme design, compared with the traditional method that relies on a large amount of manual mapping, the combination of UWB beacons and multiple sensors greatly reduces the manpower and time costs, is suitable for large-scale mine scenes, reduces the cost of manual on-site mapping and instrument measurement, and improves the applicability of large-scale underground scenes; using multiple sensors and fusion algorithms, combined with UWB beacons, laser radars, and inertial measurement units, the system can perform corrections and compensation between multiple sensor data, has stronger anti-interference and error correction capabilities, enhances the system's anti-interference ability and robustness in complex environments, and ensures the stability and reliability of positioning.

[0039] It should be appreciated that all combinations of the foregoing concepts, as well as additional concepts described in greater detail below, may be considered to be part of the inventive subject matter of the present disclosure, provided such concepts are not mutually inconsistent.

[0040] The foregoing and other aspects, embodiments and features of the present invention can be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the present invention, such as the features and / or beneficial effects of the exemplary embodiments, will be apparent from the following description or learned from the practice of the specific embodiments according to the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings are not drawn to scale according to actual reference objects. In the accompanying drawings, each identical or nearly identical component shown in various figures may be represented by the same reference numeral. For the sake of clarity, not every component is labeled in each figure. Now, embodiments of various aspects of the present invention will be described by way of example and with reference to the accompanying drawings, in which:

[0042] Figure 1 This is a flow chart of a multi-sensor fusion positioning method for a multi-layer tunnel scene according to an embodiment of the present application.

[0043] Figure 2 This is a flow chart of a method for generating a first global map of several different time periods according to an embodiment of the present application.

[0044] Figure 3This is a partial flow chart of a method for obtaining a rough position of a vehicle according to an embodiment of the present application. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used herein should be the common meaning understood by people with general skills in the field to which the present invention belongs.

[0046] The words "first", "second" and similar terms used in the patent specification and claims of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular form "a" or "an" shall not be construed as "a" or "an";

[0047] The words “a”, “an”, “the” and the like do not denote a limitation of quantity, but rather denote the presence of at least one.

[0048] The words "include" or "comprises" and the like mean that the elements or objects appearing before "includes" or "comprises" include the features, wholes, steps, operations, elements and / or components listed after "includes" or "comprises", and do not exclude the existence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0049] In the existing technology, when realizing multi-sensor data fusion positioning, on-site mapping and instrument measurement are often used to obtain the global position of the ultra-wideband beacon, that is, the global position of the UWB beacon. This method has a high labor cost and is not suitable for large-scale underground scenes. Laser radar is also used to achieve local precise positioning, but it cannot achieve startup at any position.

[0050] In view of this, an embodiment of the present invention provides a multi-sensor fusion positioning method for a multi-layer tunnel scene, referring to Figure 1 ,include:

[0051] Step S102: collecting attitude data of an inertial measurement unit located on the vehicle and used to detect the lane environment, point cloud data of a laser radar, and UWB beacon data arranged in the lane at different time periods;

[0052] Step S104: importing point cloud data and posture data into the SLAM algorithm to generate a first global map of several different time periods;

[0053] Step S106: merging a plurality of first global maps to generate a second global map;

[0054] Step S108: splitting the second global map to generate a number of topological sub-maps;

[0055] Step S110: Obtain vehicle positioning according to UWB beacon data.

[0056] If the tunnel is divided into multiple sections, the relative position of the UWB beacon in each section of the tunnel remains constant, such as height and spatial position, and each UWB beacon can be clearly scanned by the radar to improve the accuracy of subsequent labeling and positioning calculations. By integrating the data of the inertial measurement unit, lidar and UWB beacon, the accuracy and reliability of positioning can be improved. A significant feature of technological fusion is the improvement of accuracy and real-time performance. LiDAR provides rich environmental information and precise distance measurement, the inertial measurement unit provides fast motion state feedback, and UWB provides the initial position value at any position. The combination of these three enables the embodiments of the present invention to start high-frequency positioning at any position and achieve real-time high-precision positioning.

[0057] Data collection in different time periods can reflect the dynamic changes of the environment and improve the accuracy of the global map. Combining the data collection of inertial measurement units, lidars and UWB beacons to achieve multi-sensor data fusion can ensure stable vehicle positioning in complex tunnel environments. Data collection and SLAM algorithm processing in different time periods ensure the adaptability of the system in dynamic environments. A multi-sensor fusion positioning method for multi-layer tunnel scenes provided by an embodiment of the present invention is suitable for complex multi-layer tunnel scenes. It integrates data from multiple sensors such as lidars, inertial measurement units and UWBs. Through the fine processing and deep fusion of these data, a high-frequency positioning function that can be started at any location is realized.

[0058] In some embodiments, reference Figure 2 , the method of importing point cloud data and posture data into the SLAM algorithm to generate a first global map of several different time periods includes:

[0059] Step S1041: constructing several layers of point cloud maps according to the point cloud data;

[0060] Step S1042: labeling each layer of the point cloud map with a layer number according to the roadway hierarchy relationship;

[0061] Step S1043: outputting the last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud maps with adjacent layer numbers in sequence;

[0062] Step S1044: Fusing several layers of point cloud maps to generate a first global map;

[0063] Step S1045: Import the last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud map with adjacent layer numbers into the ICP algorithm to generate a first global map through point-to-surface ICP.

[0064] By integrating hierarchical relationship annotation and odometer pose, we can achieve effective fusion of point cloud maps at different levels and generate an accurate first global map. Different from the conventional point-to-point ICP method, we calculate the distance from the source point cloud to the surface composed of the target point cloud, and use the ICP algorithm to achieve point-to-surface matching, which can improve the accuracy of map stitching and reduce error accumulation. We use LiDAR point cloud data to generate multi-layer point cloud maps, and combine odometer pose with hierarchical information to ensure the global continuity and consistency of the map, thereby improving the accuracy of positioning.

[0065] In some embodiments, the method of merging several first global maps to generate a second global map includes: fusing map segments of overlapping areas in the several first global maps by using a map merging method to generate a second global map.

[0066] By fusing the map segments in the overlapping areas, the continuity and accuracy of the map can be further improved, and a complete second global map can be generated. By merging the first global map generated in multiple time periods, the problem of incomplete maps generated by a single SLAM algorithm is solved, ensuring that the global map has a wider coverage and the positioning results are more stable and accurate.

[0067] In some embodiments, the method of obtaining the precise location of a vehicle based on UWB beacon data includes:

[0068] The rough position of the vehicle is obtained based on the UWB beacon data; the rough position of the vehicle, the real-time lidar data of the vehicle, and the second global map are matched to obtain the vehicle positioning.

[0069] The UWB beacon data is used to quickly obtain the rough position of the vehicle, providing a basis for subsequent precise positioning. By matching the LiDAR data and the global map, the vehicle positioning is further refined to ensure the accuracy and reliability of positioning. Combining UWB beacons and LiDAR data, efficient conversion from coarse positioning to precise positioning is achieved. The UWB beacon provides rough location information, which is combined with the LiDAR point cloud and the global map to further improve the positioning accuracy.

[0070] In some embodiments, the method of obtaining vehicle positioning according to UWB beacon data further includes: updating the positioning of the vehicle in the topological sub-map according to the posture data of the inertial measurement unit.

[0071] Using the attitude data of the inertial measurement unit, the positioning of the vehicle in the topological sub-map can be corrected in real time, the cumulative error can be reduced, and the accuracy of dynamic positioning can be improved. In addition to the UWB beacon data, the attitude data of the inertial measurement unit is used to dynamically correct the vehicle position, compensate for the error in the sensor fusion process, and make the positioning more accurate. Due to the low frequency of the lidar, positioning drift will occur when the vehicle makes a sharp turn. At this time, the high-frequency odometer posture data is used to interpolate and update the current position of the vehicle to obtain higher-frequency positioning information, such as coordinate points and headings, for use by subsequent navigation algorithms. When facing complex environments, it is adaptable. Due to the complexity of the underground environment, maps often overlap under the same z-axis coordinate. The significance of the division of the topological map is to split some special scenes, such as maps with overlapping and interlaced maps, and switch maps for positioning according to the scene, which can reduce the positioning pressure. In the environment of multi-layer tunnels, map and positioning functions are provided to provide data support for downstream navigation tasks. When the vehicle moves to the junction of two topological maps, the second scene map is switched, and the positioning of the entire scene is completed by analogy.

[0072] In some embodiments, reference Figure 3 According to the UWB beacon data, the method for obtaining the rough position of the vehicle includes: obtaining the point P whose reflection intensity value is greater than A according to the reflection intensity of the reflective sticker posted on the UWB beacon, and the reflective sticker is provided with a label number according to the order in which the vehicles enter the lane; wherein the value of A is the critical value used to distinguish the reflective sticker from other reflective objects in the lane environment; according to the light intensity point P, the nearest neighbor search is performed using KD-Tree to obtain the n points closest to the point P, and the points whose distance from the point P is less than r are clustered in the set Q. ; Wherein, n is a natural number greater than, and r is the value of the longest straight line segment on the reflective surface of the reflective tape; if the number of elements in the set Q continues to increase, then arbitrarily select a point other than point P in the set Q as point K, and repeat the previous step until the number of elements in the set Q no longer increases; if the number of elements in the set Q no longer increases, then end the nearest neighbor search and obtain the point cloud intensity of the reflective tape; the point cloud intensity of the reflective tape, output the three-dimensional coordinates of the center point of the reflective tape and the label number of the reflective tape, and obtain the rough position of the vehicle.

[0073] By detecting the light intensity of the reflective sticker, the rough position of the vehicle can be determined quickly and reliably. The KD-Tree algorithm is used for neighbor search to improve the speed and accuracy of rough positioning. The use of reflective stickers combined with light intensity to perform preliminary vehicle position identification, and the KD-Tree algorithm is used to optimize the position search. This method improves the positioning efficiency in complex scenes where UWB beacons are arranged. The use of UWB is introduced in positioning based on point cloud maps, and the point cloud where the UWB beacon is located is clustered and output based on the point cloud intensity during the early map construction, and the coordinates and numbers in the point cloud map are output. The introduction of reflective stickers with high reflective intensity and the clustering method based on the point cloud intensity greatly improve the processing speed of the coordinates of the UWB beacon in the point cloud map, and reduce the processing workload. In extremely harsh environments, if the positioning is lost, the current vehicle position is calculated based on its own coordinates through UWB, and the positioning is calibrated, which will significantly improve the stability of the entire positioning system and its adaptability to the environment.

[0074] Another embodiment of the present invention further provides a multi-sensor fusion positioning device for a multi-layer tunnel scene, comprising:

[0075] An acquisition module, used to collect attitude data of an inertial measurement unit located on the vehicle and used to detect the roadway environment, point cloud data of a laser radar, and data of UWB beacons arranged at equal intervals in the roadway at different time periods;

[0076] A first generation module is used to import point cloud data and posture data into the SLAM algorithm to generate a first global map of several different time periods;

[0077] A second generating module, used for merging a plurality of first global maps to generate a second global map;

[0078] A third generation module is used to split the second global map to generate a plurality of topological sub-maps;

[0079] The acquisition module is used to obtain the vehicle positioning according to the UWB beacon data.

[0080] Another embodiment of the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, a multi-sensor fusion positioning method for a multi-layer tunnel scene is implemented.

[0081] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the data processing device of the gateway, and uses various interfaces and lines to connect various parts of the data processing device of the entire gateway.

[0082] Another embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, a multi-sensor fusion positioning method for a multi-layer tunnel scene is implemented.

[0083] The memory may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created by the processor, etc. In addition, the memory is preferably but not limited to a high-speed random access memory, for example, it may also be a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may also optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0084] Those skilled in the art can understand that the implementation of all or part of the processes in the above-mentioned embodiments can be completed by instructing related hardware through a computer program, which can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0085] In summary, by using a multi-sensor fusion positioning method, device, electronic device and storage medium for multi-layer tunnel scenes provided by the present invention, the inertial measurement unit, lidar and UWB beacon data are fused through multi-sensor data fusion, so that high-precision positioning can be achieved in the complex environment of multi-layer tunnels; by data collection and map generation in different time periods, and real-time updating of vehicle posture data, it is possible to adapt to the changes in tunnel structure in dynamic environments such as mines, enhance the flexibility of the system, and ensure the dynamic response capability and accuracy of the positioning system; through systematic scheme design, compared with the traditional method that relies on a large amount of manual surveying and mapping, the combination of UWB beacons and multiple sensors greatly reduces the manpower and time costs, is suitable for large-scale mine scenes, reduces the cost of manual on-site surveying and instrument measurement, and improves the applicability of large-scale underground scenes; using multiple sensors and fusion algorithms, combined with UWB beacons, lidar, and inertial measurement units, the system can perform corrections and compensation between multiple sensor data, has stronger anti-interference and error correction capabilities, enhances the system's anti-interference ability and robustness in complex environments, and ensures the stability and reliability of positioning.

[0086] In a complex multi-layer tunnel environment, a high-frequency positioning function without a specific initial position is realized. The technical solution integrates the data of laser radar, inertial measurement unit and ultra-wideband, and may add more sensors. Through the deep fusion of these data, a comprehensive understanding and precise positioning of the environment are achieved. Compared with the existing technology, the solution of the present invention significantly improves the accuracy and real-time performance of positioning by virtue of its unique sensor fusion strategy. The laser radar provides rich environmental information and accurate distance measurement, the inertial measurement unit provides fast motion state feedback, and UWB provides the initial position value of any position. The combination of these three enables the solution to start high-frequency positioning at any position and achieve real-time high-precision positioning. Especially in the environment of multi-layer tunnels, traditional positioning technologies often face serious challenges, such as signal blocking, multipath effects and other problems. However, the solution of the present invention effectively overcomes these problems by utilizing the advantages of multiple sensors and achieves high-precision positioning in such a complex environment.

[0087] Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. A person with ordinary knowledge in the technical field to which the present invention belongs may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the definition of the claims.

Claims

1. A multi-sensor fusion positioning method for multi-layer tunnel scenes, characterized in that: include: Collect attitude data of the inertial measurement unit located on the vehicle and used to detect the roadway environment, point cloud data of the laser radar, and data of the UWB beacon arranged in the roadway at different time periods; Importing the point cloud data and the posture data into a SLAM algorithm to generate a first global map of a plurality of different time periods; Merging a plurality of the first global maps to generate a second global map; splitting the second global map to generate a plurality of topological sub-maps; The vehicle positioning is obtained according to the UWB beacon data.

2. A multi-sensor fusion positioning method for multi-layer tunnel scenes according to claim 1, characterized in that: The method of importing the point cloud data and the posture data into the SLAM algorithm to generate a first global map of several different time periods includes: Constructing several layers of point cloud maps according to the point cloud data; Label each layer of the point cloud map according to the roadway hierarchy; Output the last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud maps with adjacent layer numbers in sequence; Fusing several layers of point cloud maps to generate the first global map; The last odometer pose of the previous point cloud map and the first odometer pose of the next point cloud map in the point cloud map with adjacent layer numbers are imported into the ICP algorithm to generate the first global map through point-to-surface ICP.

3. The multi-sensor fusion positioning method for multi-layer tunnel scenes according to claim 1 is characterized in that: The method of merging a plurality of the first global maps to generate a second global map includes: fusing map segments of overlapping areas in a plurality of the first global maps by a map merging method to generate the second global map.

4. The multi-sensor fusion positioning method for multi-layer tunnel scenes according to claim 1 is characterized in that: The method for obtaining the precise position of the vehicle according to the UWB beacon data includes: Obtaining a rough position of the vehicle according to the UWB beacon data; The rough position of the vehicle, the real-time lidar data of the vehicle, and the second global map are matched to obtain the vehicle positioning.

5. The multi-sensor fusion positioning method for multi-layer tunnel scenes according to claim 4 is characterized in that: The method for obtaining vehicle positioning according to the UWB beacon data further includes: updating the vehicle's positioning on the topological sub-map according to the posture data of the inertial measurement unit.

6. The multi-sensor fusion positioning method for multi-layer tunnel scenes according to claim 4 is characterized in that: The method for obtaining a rough position of a vehicle according to the UWB beacon data comprises: According to the reflection intensity of the reflective sticker posted on the UWB beacon, a point P with a reflection intensity value greater than A is obtained, wherein the reflective sticker is provided with label numbers according to the order in which the vehicles enter the lane; wherein the value of A is a critical value for distinguishing the reflective sticker from other reflective objects in the lane environment; According to the light intensity point P, a nearest neighbor search is performed using KD-Tree to obtain n points closest to the point P, and the points whose distance from the point P is less than r are clustered into a set Q; wherein n is a natural number greater than, and r is the value of the longest straight line segment on the reflective surface of the reflective tape; If the number of elements in the set Q continues to increase, then any point other than the point P is selected from the set Q as the point K, and the previous step is repeated until the number of elements in the set Q no longer increases; If the number of elements in the set Q does not increase any more, the nearest neighbor search is terminated to obtain the point cloud intensity of the reflective tape; The point cloud intensity of the reflective tape is used to output the three-dimensional coordinates of the center point of the reflective tape and the label number of the reflective tape to obtain the rough position of the vehicle.

7. A multi-sensor fusion positioning device for multi-layer tunnel scenes, characterized in that: include: An acquisition module, used to collect attitude data of an inertial measurement unit located on the vehicle and used to detect the roadway environment, point cloud data of a laser radar, and data of UWB beacons arranged at equal intervals in the roadway at different time periods; A first generating module, used for importing the point cloud data and the posture data into a SLAM algorithm to generate a first global map of a plurality of different time periods; A second generating module, used for merging a plurality of the first global maps to generate a second global map; A third generating module, used for splitting the second global map to generate a plurality of topological sub-maps; The acquisition module is used to obtain the vehicle positioning according to the UWB beacon data.

8. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, a multi-sensor fusion positioning method for a multi-layer tunnel scene as described in any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, a multi-sensor fusion positioning method for a multi-layer tunnel scene as described in any one of claims 1 to 7 is implemented.

Images