Vehicle edge cleaning trajectory generation method and system

Abstract

The present application relates to the technical field of automatic driving, and provides a vehicle edge cleaning track generation method and system, the method comprising: obtaining road design data of a target road, extracting at least one set of large-range road boundary data according to the road design data; converting the large-range road boundary data into a target format to obtain target format road boundary data; using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data; based on each set of calibrated road boundary data, generating two back-and-forth edge cleaning sparse tracks according to vehicle information; interpolating and smoothing the sparse tracks to obtain a target edge cleaning track. The present application extracts road boundary data from road design data, calibrates using laser point cloud after coordinate conversion, combines vehicle information, generates a target edge cleaning track through interpolation, and realizes lower cost, higher efficiency, and higher accuracy of edge cleaning track generation.

Description

Technical Field

[0001] This invention relates to the field of autonomous driving technology, and in particular to a method and system for generating vehicle edge-to-edge cleaning trajectories. Background Technology

[0002] As roads are continuously built, the workload of road cleaning is increasing, and it is becoming more and more difficult to recruit cleaning workers. Therefore, driverless cleaning vehicles have become an important way to solve the current problem.

[0003] In reality, nearly half of the garbage on the road is concentrated in the edge area. If unmanned sweepers cannot effectively clean along the edges, it will inevitably greatly affect their cleaning quality. Moreover, since cleaning along the edges is very easy to collide with buildings, ensuring the accuracy and safety of unmanned sweepers driving along the edges during road operations is the key to the implementation of the technology. Therefore, how to generate high-precision and high-safety edge cleaning trajectories is particularly important.

[0004] Currently, there are two common methods for generating edge-following cleaning trajectories. One method uses high-precision maps to generate edge-following driving trajectories, which the autonomous driving system then follows. The other method uses manually recorded trajectories to generate edge-following driving trajectories, which the autonomous driving system then follows. However, generating edge-following driving trajectories using high-precision maps requires the creation of large-scale high-precision maps of the road in advance, and this method requires maintaining updated road data, resulting in high costs and low efficiency. When generating edge-following driving trajectories using manually recorded trajectories, it is difficult for human drivers to maintain accurate edge-following trajectories over long distances, leading to low efficiency and low path accuracy.

[0005] In summary, existing edge-following cleaning trajectory generation methods are costly, inefficient, and lack accuracy. Summary of the Invention

[0006] This invention provides a method and system for generating vehicle edge-following cleaning trajectories, which addresses the shortcomings of existing technologies such as high cost, low efficiency, and low accuracy, and achieves edge-following cleaning trajectory generation with lower cost, higher efficiency, and higher accuracy.

[0007] This invention provides a method for generating a vehicle edge-following cleaning trajectory, comprising:

[0008] Obtain road design data for the target road, and extract at least one set of large-scale road boundary data based on the road design data;

[0009] The large-scale road boundary data is converted into a target format to obtain the target format road boundary data;

[0010] The target format road boundary data is calibrated using laser point cloud visualization to obtain calibrated road boundary data.

[0011] Based on the calibrated road boundary data of each group, two sparse cleaning trajectories along the edge are generated according to the vehicle information.

[0012] The sparse trajectory is smoothed by interpolation to obtain the target edge-following cleaning trajectory.

[0013] According to the present invention, a method for generating a vehicle edge-sweeping trajectory, wherein the method uses laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data, specifically includes:

[0014] A set of GNSS and lidar data was collected on the target road;

[0015] The GNSS and lidar data are converted into two dimensions to obtain two-dimensional boundary data;

[0016] The offset is determined by superimposing the two-dimensional boundary data and the target format road boundary data;

[0017] The target format road boundary data is adjusted according to the offset to obtain calibrated road boundary data.

[0018] According to the present invention, a method for generating vehicle edge-following cleaning trajectories includes generating two sparse edge-following cleaning trajectories in both directions based on vehicle information, according to each set of calibrated road boundary data. Specifically, this includes:

[0019] S1: Based on a set of calibration road boundary data and vehicle information, a first scaling factor and a second scaling factor are constructed. The first edge-following sparse sweeping trajectory is calculated using the first scaling factor, the second scaling factor and the calibration road boundary data.

[0020] S2: Exchange the first scaling factor and the second scaling factor, and use the exchanged first scaling factor, the second scaling factor and the calibrated road boundary data to calculate the second edge-following sparse sweeping trajectory.

[0021] According to a method for generating a vehicle edge-following cleaning trajectory provided by the present invention, the vehicle information includes the distance from the vehicle center to the road boundary; the first scaling factor includes the ratio of the distance from the vehicle center to the road boundary to the road width of the target road; and the sum of the second scaling factor and the first scaling factor is 1.

[0022] According to the method for generating a vehicle edge-sweeping trajectory provided by the present invention, the step of acquiring a set of GNSS and lidar data on the target road further includes:

[0023] Based on the vehicle-mounted lidar and GNSS equipment, a set of GNSS and lidar data is collected;

[0024] A hand-eye calibration model is constructed based on the GNSS and lidar data, and the external parameter calibration of lidar to GNSS is completed according to the hand-eye calibration model.

[0025] According to the present invention, a method for generating a vehicle edge-following cleaning trajectory, wherein the interpolation smoothing of the sparse trajectory to obtain the target edge-following cleaning trajectory specifically includes:

[0026] The sparse trajectory is smoothed by interpolation using a spiral curve smoothing algorithm to obtain the target edge-fitting cleaning trajectory.

[0027] According to the present invention, a method for generating a vehicle edge-following sweeping trajectory includes extracting at least two sets of large-scale road boundary data based on the road design data; exchanging the first scaling factor and the second scaling factor, and calculating a second edge-following sparse sweeping trajectory using the exchanged first scaling factor, the second scaling factor, and the calibrated road boundary data; and further comprising:

[0028] Repeat steps S1-S2 to generate two edge-following cleaning trajectories for the second set of boundaries.

[0029] The present invention also provides a vehicle edge-following cleaning trajectory generation system, comprising:

[0030] The acquisition module is used to acquire road design data of the target road and extract at least one set of large-scale road boundary data based on the road design data.

[0031] The format module is used to convert the large-scale road boundary data into a target format to obtain the target format road boundary data;

[0032] The calibration module is used to calibrate the target format road boundary data using laser point cloud visualization to obtain calibrated road boundary data;

[0033] The trajectory module is used to generate two back-and-forth sparse cleaning trajectories along the edge based on the vehicle information of each set of calibrated road boundary data.

[0034] The smoothing module is used to interpolate and smooth the sparse trajectory to obtain the target edge-fitting cleaning trajectory.

[0035] 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 the processor executes the program to implement the vehicle edge cleaning trajectory generation method described above.

[0036] The present invention also provides a vehicle, including a vehicle edge cleaning trajectory generation system as described in any of the above descriptions or an electronic device as described in any of the above descriptions.

[0037] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the vehicle edge cleaning trajectory generation method as described above.

[0038] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the vehicle edge cleaning trajectory generation method described above.

[0039] This invention provides a method and system for generating vehicle edge-following cleaning trajectories. The method involves acquiring road design data for a target road, extracting at least one set of large-scale road boundary data based on the road design data, converting the large-scale road boundary data into a target format, obtaining target format road boundary data, calibrating the target format road boundary data using laser point cloud visualization, obtaining calibrated road boundary data, generating two sparse edge-following cleaning trajectories for each set of calibrated road boundary data based on vehicle information, and then interpolating and smoothing the sparse trajectories to obtain the target edge-following cleaning trajectory. This invention extracts large-scale road boundary data from road design data; calibrates the large-scale road boundary data using laser point cloud after coordinate transformation; generates two sparse trajectories for each set of boundaries based on vehicle information; and then uses a smoothing algorithm to interpolate and smooth the sparse trajectories to generate the target edge-following cleaning trajectory, achieving lower cost, higher efficiency, and higher accuracy in edge-following cleaning trajectory generation. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is one of the flowcharts illustrating the vehicle edge-sweeping trajectory generation method provided by the present invention;

[0042] Figure 2 This is the second flowchart illustrating the vehicle edge-sweeping trajectory generation method provided by the present invention;

[0043] Figure 3 These are two sets of road boundary diagrams for the vehicle edge-following sweeping trajectory generation method provided by this invention;

[0044] Figure 4This is a schematic diagram of the road boundary and the warning strip boundary identified by lidar, representing the design data for the vehicle edge-sweeping trajectory generation method provided by this invention;

[0045] Figure 5 This is a schematic diagram of the sparse edge-sweeping trajectory generated by combining vehicle information in the vehicle edge-sweeping trajectory generation method provided by the present invention.

[0046] Figure 6 This is a schematic diagram of the smoothed target edge-sweeping trajectory of the vehicle edge-sweeping trajectory generation method provided by the present invention;

[0047] Figure 7 This is a schematic diagram of the vehicle edge-following cleaning trajectory generation system provided by the present invention;

[0048] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention 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 invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0050] The following is combined with Figures 1-6 The present invention describes a method for generating vehicle edge-sweeping trajectories. Figures 1-2 This is a flowchart illustrating the vehicle edge-following cleaning trajectory generation method provided by the present invention, as shown below. Figure 1 As shown, the method includes:

[0051] Step 110: Obtain the road design data of the target road, and extract at least one set of large-scale road boundary data based on the road design data.

[0052] The process involves acquiring road design data for the target road and using this data to generate large-scale road boundary data. It's important to understand that the road design data doesn't necessarily have to be high-precision design data. In fact, road design data is generally not high-precision design data. Furthermore, the large-scale road boundary data should consist of at least one set, with the left and right boundaries forming one set.

[0053] It is understood that the target road can be any road that requires the generation of vehicle edge-sweeping trajectories, such as highways and ordinary roads, and this invention does not impose any limitations on it. In one embodiment, highway design data of the target highway is obtained.

[0054] Furthermore, in the specific implementation process, the large-scale road boundary data is the road boundary data under the geodetic 2000 coordinate system.

[0055] In some embodiments, two sets of large-scale road boundary data are directly extracted, and the resulting large-scale road boundary data is illustrated in the figure below. Figure 3 As shown. The vehicle edge-sweeping trajectory generation method provided by this invention uses road design data to generate the vehicle edge-sweeping trajectory, which can be used on roads without high-precision maps and does not require the creation of high-precision maps.

[0056] Step 120: Convert the large-scale road boundary data into the target format to obtain the target format road boundary data.

[0057] The large-scale road boundary data is converted into a target format. It is understood that the vehicle edge-following cleaning trajectory generated by this invention is applied to an autonomous driving system, and the target format is a format usable by the autonomous driving system, including but not limited to UTM coordinate data format. In some embodiments, the large-scale road boundary data is geodetic 2000 coordinate system road boundary data, and the target format is UTM coordinate data.

[0058] Step 130: Use laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data.

[0059] The target format road boundary data is calibrated using laser point cloud visualization. In some embodiments, two sets of large-scale road boundary data are directly extracted, and in this step, the two sets of target format road boundary data are simultaneously calibrated using laser point cloud visualization.

[0060] Step 140: Based on the calibrated road boundary data of each group, generate two back-and-forth sparse sweeping trajectories along the edge according to the vehicle information.

[0061] Based on vehicle information and each set of calibrated road boundary data, two sparse cleaning trajectories are generated, one going back and one forth along the edge. Figure 5 As shown. In some embodiments, the vehicle information includes at least the distance from the vehicle's center to the road boundary, and the vehicle information may be, for example, information about sanitation vehicles.

[0062] Step 150: Interpolate and smooth the sparse trajectory to obtain the target edge cleaning trajectory.

[0063] In some embodiments, the interpolation smoothing of the sparse trajectory to obtain the target edge-following cleaning trajectory specifically includes:

[0064] The sparse trajectory is smoothed using a spiral curve smoothing algorithm to obtain the target edge-fitting cleaning trajectory, such as... Figure 6 As shown.

[0065] This invention eliminates the need for pre-recording driving trajectories and creating high-precision maps. It can generate long-distance, high-precision edge-following driving trajectories in one go and can generate different edge-following paths based on vehicle information, resulting in lower costs, higher efficiency, and higher accuracy.

[0066] In some embodiments, the step of using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data specifically includes:

[0067] A set of GNSS and lidar data was collected on the target road;

[0068] The GNSS and lidar data are converted into two dimensions to obtain two-dimensional boundary data;

[0069] The offset is determined by superimposing the two-dimensional boundary data and the target format road boundary data;

[0070] The target format road boundary data is adjusted according to the offset to obtain calibrated road boundary data.

[0071] Specifically, such as Figure 4 As shown, a set of GNSS (Global Navigation Satellite System) and lidar data is collected on the road under construction (target road); then the lidar data is 2D and two sets of road boundary data are superimposed to determine the offset; the two sets of road boundary data are then adjusted.

[0072] In some embodiments, since the road boundary consists of a series of coordinate points, when adjusting each set of road boundary data, each set of road boundary data is divided into a series of vectors by dividing each pair of coordinate points into a vector. Then, the vector translation method is used to process each vector in turn, translating it in the direction perpendicular to the vector by the offset determined in the previous step. Then, all the processed vectors are reassembled into a single set of road boundary data to complete the adjustment of the set of road boundary data.

[0073] It is important to note that the data acquired by GNSS and lidar are, in fact, the boundaries of the warning strip. In one embodiment, translation is used to align the warning strip boundary represented by the two-dimensional boundary data with the target-format road boundary data, thus determining the offset.

[0074] In some embodiments, the acquisition of a set of GNSS and lidar data on the target road may be preceded by:

[0075] Based on the vehicle-mounted lidar and GNSS equipment, a set of GNSS and lidar data is collected;

[0076] A hand-eye calibration model is constructed based on the GNSS and lidar data, and the external parameter calibration of lidar to GNSS is completed according to the hand-eye calibration model.

[0077] Specifically, in actual operation, firstly, lidar and GNSS equipment are installed on the vehicle. Based on the lidar and GNSS equipment installed on the vehicle, a set of GNSS and lidar data is collected on the target road. Based on the collected GNSS and lidar data, a hand-eye calibration model is constructed to complete the calibration of the external parameters (transformation matrix) from lidar to GNSS.

[0078] In some embodiments, the step of generating two sparse, edge-following cleaning trajectories based on vehicle information, according to each set of calibrated road boundary data, specifically includes:

[0079] S1: Based on a set of calibration road boundary data and vehicle information, a first scaling factor and a second scaling factor are constructed. The first edge-following sparse sweeping trajectory is calculated using the first scaling factor, the second scaling factor and the calibration road boundary data.

[0080] S2: Exchange the first scaling factor and the second scaling factor, and use the exchanged first scaling factor, the second scaling factor and the calibrated road boundary data to calculate the second edge-following sparse sweeping trajectory.

[0081] Specifically, for a set of calibrated road boundary data, a first edge-following sparse sweeping trajectory is generated based on vehicle information using a pre-built trajectory generation formula; then, by exchanging the scaling factor, a second edge-following sweeping trajectory is generated based on vehicle information using the pre-built trajectory generation formula.

[0082] In some embodiments, the vehicle information includes the distance from the vehicle center to the road boundary; the first scaling factor includes the ratio of the distance from the vehicle center to the road boundary to the road width of the target road; and the sum of the second scaling factor and the first scaling factor is 1.

[0083] In some embodiments, a first edge-following sparse sweeping trajectory is calculated using a trajectory formula based on a first scaling factor, a second scaling factor, and calibrated road boundary data. The trajectory generation formula includes:

[0084] L = ratio·B l +(1-ratio)·B r

[0085] Where L represents the sparse trajectory of the edge-to-edge cleaning; ratio is the first scaling factor, ratio = distance from the vehicle center to the road boundary / road width; (1-ratio) is the second scaling factor; B lB represents the coordinates of the left boundary of the lane. l These are the coordinates of the right boundary of the lane.

[0086] In some embodiments, at least two sets of large-scale road boundary data are extracted based on the road design data; after exchanging the first scaling factor and the second scaling factor, and using the exchanged first scaling factor, the second scaling factor, and the calibrated road boundary data to calculate a second edge-following sparse sweep trajectory, the method further includes:

[0087] Repeat steps S1-S2 to generate two edge-following cleaning trajectories for the second set of boundaries.

[0088] Specifically, when extracting at least two sets of large-scale road boundary data based on road design data, two back-and-forth sparse cleaning trajectories are generated based on vehicle information, sequentially using each set of calibrated road boundary data. The specific steps are as follows:

[0089] For the first set of large-scale road boundary data, according to the formula ratio·B l +(1-ratio)·B r Generate the first edge-following cleaning trajectory (ratio = distance from vehicle center to road boundary / road width, B). l B r (Coordinates of the left and right boundaries of the lane);

[0090] Then, the scaling factor (ratio = 1.0 - ratio) is swapped to generate a second edge-following cleaning trajectory;

[0091] For the second set of large-scale road boundary data, repeat steps S1 and S2 to generate two more edge-following cleaning trajectories.

[0092] This invention provides a method for generating vehicle edge-following cleaning trajectories. The method involves acquiring road design data for a target road, extracting at least one set of large-scale road boundary data based on the road design data, converting the large-scale road boundary data into a target format, obtaining target format road boundary data, calibrating the target format road boundary data using laser point cloud visualization, obtaining calibrated road boundary data, generating two sparse edge-following cleaning trajectories for each set of calibrated road boundary data based on vehicle information, and then interpolating and smoothing the sparse trajectories to obtain the target edge-following cleaning trajectory. This invention extracts large-scale road boundary data from road design data; calibrates the large-scale road boundary data using laser point cloud after coordinate transformation; generates two sparse trajectories for each set of boundaries based on vehicle information; and then uses a smoothing algorithm to interpolate and smooth the sparse trajectories to generate the target edge-following cleaning trajectory, achieving lower cost, higher efficiency, and higher accuracy in edge-following cleaning trajectory generation.

[0093] The vehicle edge cleaning trajectory generation system provided by the present invention is described below. The vehicle edge cleaning trajectory generation system described below can be referred to in correspondence with the vehicle edge cleaning trajectory generation method described above. Figure 7 This is a schematic diagram of the vehicle edge-following cleaning trajectory generation system provided by the present invention. (See diagram below.) Figure 7 As shown, the system includes:

[0094] The acquisition module 710 is used to acquire road design data of the target road and extract at least one set of large-scale road boundary data based on the road design data.

[0095] Format module 720 is used to convert the large-scale road boundary data into a target format to obtain target format road boundary data;

[0096] Calibration module 730 is used to calibrate the target format road boundary data using laser point cloud visualization to obtain calibrated road boundary data;

[0097] The trajectory module 740 is used to generate two back-and-forth sparse cleaning trajectories along the edge based on the vehicle information, according to the calibration road boundary data of each group.

[0098] The smoothing module 750 is used to interpolate and smooth the sparse trajectory to obtain the target edge cleaning trajectory.

[0099] In some embodiments, the calibration module 730 specifically includes:

[0100] A set of GNSS and lidar data was collected on the target road;

[0101] The GNSS and lidar data are converted into two dimensions to obtain two-dimensional boundary data;

[0102] The offset is determined by superimposing the two-dimensional boundary data and the target format road boundary data;

[0103] The target format road boundary data is adjusted according to the offset to obtain calibrated road boundary data.

[0104] In some embodiments, the trajectory module 740 specifically includes:

[0105] S1: Based on a set of calibration road boundary data and vehicle information, a first scaling factor and a second scaling factor are constructed. The first edge-following sparse sweeping trajectory is calculated using the first scaling factor, the second scaling factor and the calibration road boundary data.

[0106] S2: Exchange the first scaling factor and the second scaling factor, and use the exchanged first scaling factor, the second scaling factor and the calibrated road boundary data to calculate the second edge-following sparse sweeping trajectory.

[0107] In some embodiments, the vehicle information includes the distance from the vehicle center to the road boundary; the first scaling factor includes the ratio of the distance from the vehicle center to the road boundary to the road width of the target road; and the sum of the second scaling factor and the first scaling factor is 1.

[0108] In some embodiments, the acquisition of a set of GNSS and lidar data on the target road may be preceded by:

[0109] Based on the vehicle-mounted lidar and GNSS equipment, a set of GNSS and lidar data is collected;

[0110] A hand-eye calibration model is constructed based on the GNSS and lidar data, and the external parameter calibration of lidar to GNSS is completed according to the hand-eye calibration model.

[0111] In some embodiments, the smoothing module 750 specifically includes:

[0112] The sparse trajectory is smoothed by interpolation using a spiral curve smoothing algorithm to obtain the target edge-fitting cleaning trajectory.

[0113] In some embodiments, at least two sets of large-scale road boundary data are extracted based on the road design data; after exchanging the first scaling factor and the second scaling factor, and using the exchanged first scaling factor, the second scaling factor, and the calibrated road boundary data to calculate a second edge-following sparse sweep trajectory, the method further includes:

[0114] Repeat steps S1-S2 to generate two edge-following cleaning trajectories for the second set of boundaries.

[0115] This invention provides a vehicle edge-following cleaning trajectory generation system. It acquires road design data of a target road and extracts at least one set of large-scale road boundary data based on the road design data. The large-scale road boundary data is then converted into a target format to obtain target format road boundary data. The target format road boundary data is then visualized and calibrated using laser point clouds to obtain calibrated road boundary data. Based on each set of calibrated road boundary data, two sparse edge-following cleaning trajectories are generated for each direction based on vehicle information. The sparse trajectories are then interpolated and smoothed to obtain the target edge-following cleaning trajectory. This invention extracts large-scale road boundary data from road design data; calibrates the large-scale road boundary data using laser point clouds after coordinate transformation; generates two sparse trajectories for each set of boundaries based on vehicle information; and then uses a smoothing algorithm to interpolate and smooth the sparse trajectories to generate the target edge-following cleaning trajectory, achieving lower cost, higher efficiency, and higher accuracy in edge-following cleaning trajectory generation.

[0116] The present invention also provides a vehicle, including the vehicle edge cleaning trajectory generation system or electronic device as described in any of the above embodiments.

[0117] In this embodiment, the vehicle is a vehicle equipped with a vehicle-to-side cleaning function, such as a sanitation vehicle or other engineering vehicle.

[0118] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communication interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communication interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute a method for generating a vehicle edge-sweeping trajectory. This method includes: acquiring road design data of the target road; extracting at least one set of large-scale road boundary data based on the road design data; converting the large-scale road boundary data into a target format to obtain target format road boundary data; using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data; generating two sparse edge-sweeping trajectories in both directions based on each set of calibrated road boundary data and vehicle information; and interpolating and smoothing the sparse trajectories to obtain the target edge-sweeping trajectory.

[0119] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0120] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the vehicle edge-sweeping trajectory generation method provided by the above methods. The method includes: acquiring road design data of a target road; extracting at least one set of large-scale road boundary data based on the road design data; converting the large-scale road boundary data into a target format to obtain target format road boundary data; using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data; generating two back-and-forth edge-sweeping sparse trajectories based on vehicle information according to each set of calibrated road boundary data; and interpolating and smoothing the sparse trajectories to obtain the target edge-sweeping trajectory.

[0121] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements a method for generating vehicle edge-sweeping trajectories provided by the methods described above. The method includes: acquiring road design data of a target road; extracting at least one set of large-scale road boundary data based on the road design data; converting the large-scale road boundary data into a target format to obtain target format road boundary data; using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data; generating two sparse edge-sweeping trajectories in both directions based on vehicle information for each set of calibrated road boundary data; and interpolating and smoothing the sparse trajectories to obtain the target edge-sweeping trajectory.

[0122] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0123] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0124] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for generating a vehicle edge-sweeping trajectory, characterized in that, include: Obtain road design data for the target road, and extract at least one set of large-scale road boundary data based on the road design data; The large-scale road boundary data is converted into a target format to obtain the target format road boundary data; The target format road boundary data is calibrated using laser point cloud visualization to obtain calibrated road boundary data. Based on the calibrated road boundary data for each set, two sparse cleaning trajectories are generated back and forth according to vehicle information; specifically, generating two sparse cleaning trajectories back and forth according to vehicle information based on the calibrated road boundary data for each set includes: S1: Based on a set of calibration road boundary data and vehicle information, a first scaling factor and a second scaling factor are constructed. The first edge-following sparse sweeping trajectory is calculated using the first scaling factor, the second scaling factor and the calibration road boundary data. S2: Exchange the first scaling factor and the second scaling factor, and calculate the second edge-following sparse sweeping trajectory using the exchanged first scaling factor, second scaling factor, and the calibrated road boundary data; wherein, the vehicle information includes the distance from the vehicle center to the road boundary; the first scaling factor includes the ratio of the distance from the vehicle center to the road boundary to the road width of the target road; the sum of the second scaling factor and the first scaling factor is 1; The sparse trajectory is smoothed by interpolation to obtain the target edge-following cleaning trajectory.

2. The method for generating vehicle edge-sweeping trajectories according to claim 1, characterized in that, The step of using laser point cloud visualization to calibrate the target format road boundary data to obtain calibrated road boundary data specifically includes: A set of GNSS and lidar data was collected on the target road; The GNSS and lidar data are converted into two dimensions to obtain two-dimensional boundary data; The offset is determined by superimposing the two-dimensional boundary data and the target format road boundary data; The target format road boundary data is adjusted according to the offset to obtain calibrated road boundary data.

3. The method for generating vehicle edge-sweeping trajectories according to claim 2, characterized in that, The acquisition of a set of GNSS and lidar data on the target road, prior to which includes: Based on the vehicle-mounted lidar and GNSS equipment, a set of GNSS and lidar data is collected; A hand-eye calibration model is constructed based on the GNSS and lidar data, and the external parameter calibration of lidar to GNSS is completed according to the hand-eye calibration model.

4. The method for generating vehicle edge-sweeping trajectories according to claim 1, characterized in that, The interpolation smoothing of the sparse trajectory yields the target edge-following cleaning trajectory, specifically including: The sparse trajectory is smoothed by interpolation using a spiral curve smoothing algorithm to obtain the target edge-fitting cleaning trajectory.

5. The method for generating vehicle edge-sweeping trajectories according to claim 1, characterized in that, At least two sets of large-scale road boundary data are extracted based on the road design data; the first and second scaling factors are exchanged, and a second edge-following sparse trajectory is calculated using the exchanged first and second scaling factors and the calibrated road boundary data, followed by: Repeat steps S1-S2 to generate two edge-following cleaning trajectories for the second set of boundaries.

6. A vehicle edge-sweeping trajectory generation system applied to the vehicle edge-sweeping trajectory generation method as described in any one of claims 1 to 5, characterized in that, include: The acquisition module is used to acquire road design data of the target road and extract at least one set of large-scale road boundary data based on the road design data. The format module is used to convert the large-scale road boundary data into a target format to obtain the target format road boundary data; The calibration module is used to calibrate the target format road boundary data using laser point cloud visualization to obtain calibrated road boundary data; The trajectory module is used to generate two back-and-forth sparse cleaning trajectories along the edge based on the vehicle information of each set of calibrated road boundary data. The smoothing module is used to interpolate and smooth the sparse trajectory to obtain the target edge-fitting cleaning trajectory.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the vehicle edge cleaning trajectory generation method as described in any one of claims 1 to 5.

8. A vehicle, characterized in that, This includes the vehicle edge-following cleaning trajectory generation system as described in claim 6 or the electronic device as described in claim 7.

Images