A target trajectory prediction method, device, equipment, vehicle and storage medium

Abstract

The present invention belongs to the technical field of vehicle trajectory prediction, and specifically relates to a target trajectory prediction method, device, equipment, vehicle, and storage medium. A target trajectory prediction device includes a positioning module, a signal acquisition module, and a processor. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The dynamic search range is obtained in real time according to the vehicle driving state, and combined with the preset road section breadth-first search algorithm, the vehicle lanes that the measured vehicle may travel through are obtained. Then, rendering is performed, and grids are divided to output dense candidate points of the possible road sections. By monitoring the driving state information of the measured vehicle and the road, the search range is adjusted in real time, reducing the generation of a large amount of redundant data and saving computing power. Since the search range is adjusted in real time according to the driving state information of the measured vehicle, the trajectory prediction accuracy is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle trajectory prediction, and particularly relates to a target trajectory prediction method, device, equipment, vehicle and storage medium. Background Art

[0002] Currently, the technology of autonomous driving vehicles has entered a rapid development stage. Predicting the possible driving trajectories of surrounding vehicles is an ability that autonomous driving vehicles should possess, so that autonomous driving vehicles can take safe and effective actions.

[0003] At present, vehicle trajectory prediction methods are mainly implemented based on target-driven or data-driven. For the target-driven trajectory prediction method, first, a graph neural network is used to encode the high-precision map and vehicle information to obtain the global features of the vehicle to be predicted. Candidate points are sampled according to the high-precision map data to estimate the distribution of candidate targets in a given scenario. Then, motion estimation conditional on the target is performed to generate trajectories reaching each target. Finally, the hypothetical trajectories are evaluated and ranked, and the final prediction trajectory set with the top ranking is output.

[0004] The patent "Driving Trajectory Prediction Method, Vehicle and Computer Readable Storage Medium" with the Chinese patent application number CN202210057612.9 proposes a driving trajectory prediction method. By obtaining the first sampling points of a driving target within a preset period, fitting the first sampling points to obtain a target trajectory curve; obtaining a preset number of road acquisition points for fitting to obtain a road trajectory curve; obtaining vehicle driving data to obtain a driving trajectory curve; predicting a target driving trajectory curve according to the target trajectory curve, the road trajectory curve and the driving trajectory curve. However, the method of this patent does not fully utilize the road information extraction, thus affecting the final trajectory prediction accuracy.

[0005] Currently, the sampling methods of target prediction candidate points mainly adopt lane centerline interval sampling and grid uniform sampling methods. These methods often ignore the specific state information and road information of the vehicle, easily leading to a too large sampling area range and negative situations such as sampling in non-drivable areas, resulting in unreasonable trajectories when performing motion estimation with target points subsequently, and at the same time causing a large amount of unnecessary redundant data to be processed additionally, not only causing waste of computing resources, but also the candidate points obtained from the unreasonable sampling area will affect the accuracy of trajectory prediction. Summary of the Invention

[0006] The purpose of the present invention is to provide a target trajectory prediction method, device, equipment, vehicle and storage medium, which make full use of the specific state information and road information of the vehicle, reduce the generation of redundant data, and improve the accuracy of the generated trajectories.

[0007] To achieve the above technical objectives, the technical solution adopted by the present invention is as follows:

[0008] In a first aspect, an embodiment of the present application provides a target trajectory prediction method, which is applied to a target trajectory prediction device. The device includes a positioning module, a signal acquisition module, and a processor. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The method includes:

[0009] The positioning module establishes coordinates with the ground as the reference system. The signal acquisition module acquires the state information of the vehicle under test and sends the state information of the vehicle under test to the processor;

[0010] The processor calculates the state information of the vehicle under test to determine the central position P of the vehicle under test and determines the lane section where the vehicle under test is currently located;

[0011] According to a preset search radius algorithm, in combination with a preset prediction time T and the state information of the vehicle under test, a dynamic search range is determined;

[0012] According to the lane section information where the vehicle under test is currently located and the dynamic search range, through a preset breadth-first search algorithm for sections, the possible lane sections for the vehicle under test to travel are predicted and divided;

[0013] The predicted and divided lane sections are rendered to obtain the lane polygons of the possible lane sections for the vehicle under test to travel;

[0014] The lane polygons are rasterized, and the center points of all rasters are taken as candidate points, and dense candidate points of the possible sections are output.

[0015] In combination with the first aspect, in some alternative embodiments, the signal acquisition module acquires the state information of the vehicle under test, including:

[0016] The signal acquisition module acquires the body length positioning coordinates, body width coordinates, traveling speed V, acceleration A of the vehicle under test, and the heading H of the vehicle under test.

[0017] In combination with the first aspect, in some alternative embodiments, the processor calculates the state information of the vehicle under test to determine the central position P of the vehicle under test and determines the lane section where the vehicle under test is currently located, including:

[0018] A rectangular diagonal is constructed according to the body length positioning coordinates and the body width positioning coordinates, and the coordinates of the intersection point of the rectangular diagonal are the central position P of the vehicle under test;

[0019] The signal acquisition module collects the array coordinates of the center lines of multiple vehicle lane segments near the vehicle to be measured, calculates the Euclidean distances from the center lines of each vehicle lane segment to the central position P, and selects the vehicle lane segment where the center line of the vehicle lane closest to the central position P is located as the vehicle lane segment where the vehicle to be measured is located.

[0020] Combined with the first aspect, in some alternative embodiments, according to a preset search radius algorithm, in combination with a preset prediction time T and the state information of the vehicle to be measured, the dynamic search range is determined, including:

[0021] Perform weighted calculation based on the driving speed V, acceleration A, and the preset prediction time T to obtain the calculated search radius R dy ;

[0022] The processor is pre-set with a minimum search radius R0, and the calculated search radius R dy is compared with the minimum search radius R0;

[0023] Take the maximum value as the actual search radius R, generate a semi-circular dynamic search range with the central position P as the center and the actual search radius R as the radius, and the heading H coincides with the angular bisector of the dynamic search range.

[0024] Combined with the first aspect, in some alternative embodiments, according to the vehicle lane segment information where the vehicle to be measured is currently located and the dynamic search range, through a preset road segment breadth-first search algorithm, the possible vehicle lane segments for the vehicle to be measured to travel are predicted and divided, including:

[0025] During driving, the processor sets different vehicle lane segment ids for the vehicle lane segment where the vehicle to be measured is currently located and the vehicle lane segments involved in the dynamic search range;

[0026] The processor is provided with a queue Q and a list L. The queue Q is used to store the id of the vehicle lane segment where the vehicle to be measured is currently located and the ids of the adjacent road segments of the vehicle lane segment where the vehicle to be measured is currently located. The list L is used to store the ids of the possible vehicle lane segments that have been searched;

[0027] The id of the vehicle lane segment where the vehicle to be measured is currently located is added to the queue Q and the list L respectively as the starting road segment id;

[0028] During the forward movement of the vehicle to be measured, the processor sequentially adds the id of the road segment where the vehicle to be measured is currently traveling and the id of the critical road segment to the queue Q;

[0029] Traverse the queue Q, extract the lane segment id at the head of the queue Q, calculate the Euclidean distance from the center line of the lane segment at the head of the queue to the center position P, and compare it with the actual search range R;

[0030] If the Euclidean distance from the center line of the lane segment at the head of the queue to the center position P exceeds the actual search range R, pop the lane segment id at the head of the queue from the queue Q;

[0031] If the Euclidean distance from the center line of the lane segment at the head of the queue to the center position P does not exceed the actual search range R, add the lane segment id at the head of the queue to the list L, and then pop the lane segment id at the head of the queue from the queue Q;

[0032] When the elements in the queue Q are empty, output a list L containing all possible lane segments that the vehicle under test can travel.

[0033] Combined with the first aspect, in some alternative embodiments, render the lane segments divided by the prediction to obtain the road segment polygons of the lane segments that the vehicle under test may travel, including:

[0034] Render the lane segments corresponding to all lane segment ids in the list L, and output the road segment polygons of the lane segments that the vehicle under test may travel.

[0035] In a second aspect, an embodiment of the present application further provides a target trajectory prediction device, which is applied to a target trajectory prediction device. The device includes a positioning module, a signal acquisition module, and a processor. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The device includes:

[0036] An information collection unit: used to collect the vehicle state information and the center line position information of the lane segment during the travel of the vehicle under test, and perform coordinate system conversion and unification;

[0037] An analysis and processing unit: used to analyze and process the collected information, execute the breadth-first search algorithm for road segments, and determine the road segment polygons of the lane segments that the vehicle under test may travel.

[0038] In a third aspect, an embodiment of the present application further provides a target trajectory prediction device, including a positioning module, a signal acquisition module, a processor, and a storage module. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The storage module stores a computer program. When the computer program is executed by the processor, the target trajectory prediction device executes the above method.

[0039] Fourthly, an embodiment of the present application further provides a vehicle, which includes a vehicle body and the above-mentioned target trajectory prediction device, and the target trajectory prediction device is arranged on the vehicle.

[0040] Fifthly, an embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. When the computer program runs on a computer, the computer is enabled to execute the above-mentioned method.

[0041] The invention adopting the above technical solution has the following advantages:

[0042] The processor performs weighted calculation according to the collected driving speed V, acceleration A of the vehicle to be measured and the preset prediction time T, obtains the dynamically searched range in real time according to the driving state of the vehicle, combines the preset road section breadth-first search algorithm, obtains the vehicle lane section where the vehicle to be measured may travel, and then performs rendering, divides the grid and outputs the dense candidate points of the possible section. By monitoring the driving state information of the vehicle to be measured and the road, the search range is adjusted in real time, a large amount of redundant data is reduced, and the computing power is saved. Since the search range will be adjusted in real time according to the driving state information of the vehicle to be measured, the accuracy of trajectory prediction is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The present invention can be further described by the non-limiting embodiments given by the drawings;

[0044] Figure 1 It is a block diagram of the target trajectory prediction device provided by an embodiment of the present application;

[0045] Figure 2 It is a schematic flow chart of the target trajectory prediction method provided by an embodiment of the present application;

[0046] Figure 3 It is a schematic diagram of determining the vehicle lane section where the vehicle to be measured is located by the target trajectory prediction method provided by an embodiment of the present application;

[0047] Figure 4 It is a flow chart of the road section breadth-first search algorithm of the target trajectory prediction method provided by an embodiment of the present application;

[0048] Figure 5 It is a schematic diagram of the vehicle lane section of the road section breadth-first search algorithm of the target trajectory prediction method provided by an embodiment of the present application;

[0049] Figure 6 It is a block diagram of the target trajectory prediction device provided by an embodiment of the present application.

[0050] Icons: 10. Target trajectory prediction device; 11. Positioning module; 12. Signal acquisition module; 13. Processor; 200. Target trajectory prediction device; 210. Information collection unit; 220. Analysis and processing unit. Detailed implementation manners

[0051] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that in the description of the drawings or the specification, similar or identical parts are denoted by the same reference numerals, and the implementation manners not shown or described in the drawings are the forms known to those of ordinary skill in the art. In addition, the directional terms mentioned in the embodiments, such as "upper", "lower", "top", "bottom", "left", "right", "front", "rear", etc., are only references to the directions of the drawings and are not used to limit the protection scope of the present invention.

[0052] As Figure 1 shown, an embodiment of the present application provides a target trajectory prediction device 10, and the target trajectory prediction device 10 may include a positioning module 11, a signal acquisition module 12, a processor 13, and a storage module. The processor 13 is used to process the data collected by the signal acquisition module 12 to improve the accuracy of trajectory prediction.

[0053] Among them, the signal acquisition module 12 is used to collect the vehicle state information and road information of the vehicle being measured during the driving process. The processor 13 analyzes and compares the vehicle state information and road information to determine the current lane section where the vehicle being measured is located. According to the preset search radius algorithm, combined with the driving speed V, acceleration A of the vehicle being measured, and the preset prediction time T, a weighted calculation is performed to obtain the dynamically searched range in real time according to the driving state of the vehicle. Further, in combination with the preset section breadth-first search algorithm, the possible lane sections that the vehicle being measured may travel are obtained, and then rendering is performed to divide the grid and output the dense candidate points of the possible sections. By monitoring the driving state information and road of the vehicle being measured and adjusting the search range in real time, a large amount of redundant data is reduced, and the computer computing power is saved. Since the search range will be adjusted in real time according to the driving state information of the vehicle being measured, the accuracy of trajectory prediction is improved.

[0054] In this embodiment, the target trajectory prediction device 10 may be deployed on the vehicle to make the driving trajectory of the vehicle safer and more reliable. The signal acquisition module 12 and the processor 13 may be connected to the storage module, and data or programs stored in the storage module may be obtained from the storage module. Alternatively, the signal acquisition module 12 and the processor 13 are integrated with a storage module and have a data storage function.

[0055] The storage module stores a computer program, and when the computer program is executed by the processor 13, the target trajectory prediction device 10 can execute the corresponding steps in the following target trajectory prediction method.

[0056] As shown in Figure 2 the figure, the present application further provides a target trajectory prediction method, wherein the target trajectory prediction method may include: the positioning module 11 establishes coordinates with the ground as the reference system, and the signal acquisition module 12 acquires the state information of the vehicle to be measured and sends the state information of the vehicle to be measured to the processor 13;

[0057] The processor 13 calculates the state information of the vehicle to be measured, determines the central position P of the vehicle to be measured, and determines the lane section where the vehicle to be measured is currently located;

[0058] According to a preset search radius algorithm, in combination with a preset prediction time T and the state information of the vehicle to be measured, the dynamic search range is determined;

[0059] According to the lane section information where the vehicle to be measured is currently located and the dynamic search range, through a preset lane breadth-first search algorithm, the lane sections that the vehicle to be measured may travel through are predicted and divided;

[0060] The predicted and divided lane sections are rendered to obtain the lane polygon of the lane sections that the vehicle to be measured may travel through;

[0061] The lane polygon is rasterized, and the center points of all rasters are taken as candidate points, and the dense candidate points of the possible lane sections are output.

[0062] In this embodiment, the signal acquisition module 12 acquires the vehicle state information and road information of the vehicle to be measured during driving. The processor 13 respectively uses a preset search radius algorithm and a lane breadth-first search algorithm, combines the vehicle state information and road information, obtains the possible lane sections that the dynamic vehicle to be measured may travel through, performs rendering, grid division, and center point taking operations, and outputs the dense candidate points of the possible lane sections, improving the prediction accuracy.

[0063] As an optional implementation manner, the signal acquisition module 12 acquiring the state information of the vehicle to be measured includes:

[0064] The signal acquisition module 12 acquires the body length positioning coordinates, body width coordinates, driving speed V, acceleration A of the vehicle to be measured, and the heading H of the vehicle to be measured.

[0065] In this embodiment, in order to avoid collecting unnecessary data, causing data redundancy and wasting computing power, only the body length positioning coordinates, body width coordinates, driving speed V, acceleration A of the vehicle body, and the heading H of the vehicle to be measured are collected.

[0066] As Figure 3As shown, as an alternative embodiment, the processor 13 calculates the state information of the vehicle under test to determine the central position P of the vehicle under test, and determines the lane segment where the vehicle under test is currently located, including:

[0067] Construct a rectangle diagonal based on the body length positioning coordinates and the body width positioning coordinates, and the coordinates of the intersection point of the rectangle diagonal are the central position P of the vehicle under test;

[0068] The signal acquisition module 12 acquires the coordinate arrays of the center lines of multiple lane segments near the vehicle under test, calculates the Euclidean distance from each lane segment center line to the central position P, and selects the lane segment where the center line of the lane segment closest to the central position P is located as the lane segment where the vehicle under test is located.

[0069] It can be understood that a coordinate system is established with the central position of the vehicle to be predicted as the origin to determine the coordinate position of the first point on the center line of the road segment. The calculation method of the Euclidean distance from the center line of the lane segment to the central position P is:

[0070]

[0071] In this embodiment, the Euclidean distances from multiple lane segment center lines to the central position P are calculated, the obtained calculation results are compared, and the lane segment where the center line corresponding to the minimum value is located is taken as the lane segment where the vehicle under test is located.

[0072] As an alternative embodiment, according to a preset search radius algorithm, in combination with a preset prediction time T and the state information of the vehicle under test, a dynamic search range is determined, including:

[0073] Perform weighted calculation based on the driving speed V, acceleration A, and the preset prediction time T to obtain the calculated search radius R dy ;

[0074] The processor 13 pre-stores a minimum search radius R0, and compares the calculated search radius R dy with the minimum search radius R0;

[0075] Take the maximum value as the actual search radius R, generate a semi-circular dynamic search range with the central position P as the center and the actual search radius R as the radius, and the heading H coincides with the angular bisector of the dynamic search range.

[0076] In this embodiment, during the process of the vehicle under test moving on the road, the driving state of the vehicle under test will be affected by the road conditions, and the driving state information of the vehicle under test will also be affected accordingly. The search range required for generating the predicted trajectory should also change accordingly to avoid collecting useless data, which not only causes data redundancy but also affects the accuracy of the predicted trajectory. Therefore, through the search radius algorithm, the radius of the search range is set to the calculated search radius R that changes with the driving state of the vehicle under test dy ,

[0077] R dy =(αV + βA 3 )T

[0078] where α and β are constants. Under the conditions that the driving speed V of the vehicle is greater, the acceleration A is greater, and the prediction time T is longer, in order to ensure the driving safety of the vehicle under test, the radius of the search range for the vehicle under test should also be larger accordingly.

[0079] It can be understood that under the conditions of starting or traffic jams, the driving speed V and acceleration A of the vehicle at this time are not very large, which may cause the search range to be too small. Therefore, a minimum search radius R0 is preset in the processor, and the minimum search radius R0 is compared with the calculated search radius R dy , and the maximum value of them is taken as the actual search radius R,

[0080] R = max(R0, R dy )

[0081] A search range is established with the actual search radius R to ensure the safety of the predicted trajectory of the vehicle under test.

[0082] As Figure 4 and Figure 5 shown, as an optional implementation manner, according to the information of the lane section where the vehicle under test is currently located and the dynamic search range, through the preset lane breadth-first search algorithm, the lane sections that the vehicle under test may travel through are predicted and divided, including;

[0083] During the driving process, the processor 13 sets different lane section ids for the lane section where the vehicle under test is currently located and the lane sections involved in the dynamic search range;

[0084] A queue Q and a list L are set in the processor 13. The queue Q is used to store the lane section id where the vehicle under test is currently located and the ids of the adjacent lane sections of the lane section where the vehicle under test is currently located, and the list L is used to store the ids of the lane sections that may have been searched and traveled through;

[0085] The lane section ID where the vehicle under test is currently located is added to the queue Q and the list L respectively as the starting section ID;

[0086] During the forward movement of the vehicle under test, the processor sequentially adds the ID of the section where the vehicle under test is traveling and the ID of the critical section to the queue Q;

[0087] Traverse the queue Q cyclically, extract the lane section ID at the head of the queue Q, calculate the Euclidean distance from the center line of the lane section at the head of the queue to the center position P, and compare it with the actual search range R;

[0088] If the Euclidean distance from the center line of the lane section at the head of the queue to the center position P exceeds the actual search range R, pop the lane section ID at the head of the queue from the queue Q;

[0089] If the Euclidean distance from the center line of the lane section at the head of the queue to the center position P does not exceed the actual search range R, add the lane section ID at the head of the queue to the list L, and then pop the lane section ID at the head of the queue from the queue Q;

[0090] When the elements in the queue Q are empty, output a list L containing all possible lane sections that the vehicle under test can travel through.

[0091] In this embodiment, each lane section is marked with an ID. According to the above steps, determine the ID of the current lane section where the vehicle is located, add the ID of the current lane section where the vehicle is located and the ID of the adjacent section of the current lane section to the queue Q, and add the ID of the current lane section as the starting section ID to the queue Q and the list L respectively.

[0092] During the forward movement of the vehicle under test, the processor sequentially adds the ID of the section where the vehicle under test is traveling and the ID of the critical section to the queue Q;

[0093] Traverse and read the ID element at the head of the queue Q in sequence, calculate the Euclidean distance between the center line of the lane section corresponding to the ID element and the center position P, and compare the calculated result with the actual search range R;

[0094] If it does not exceed the actual search range R, it is possible for the vehicle under test to enter this lane section. Add this ID element to the list L and pop the ID element at the head of the queue Q; if it exceeds the actual search range R, the vehicle under test will not enter this lane section, and directly pop the ID element at the head of the queue Q.

[0095] After that, read the ID element at the head of the new queue Q and repeat the above steps until the elements in the queue Q are empty, and output a list L containing all possible lane sections that the vehicle under test can travel through.

[0096] It should be noted that the adjacent lane section referred to here refers to a lane section on which the vehicle under test can travel without violating traffic regulations in the current lane section.

[0097] As an optional implementation, rendering the predicted divided road segment to obtain the road segment polygon of the road segment on which the tested vehicle may travel includes:

[0098] The lane segments corresponding to all lane segment ids in the list L are rendered, and the lane segment polygons of the lane segments on which the detected vehicle may travel are output.

[0099] In this embodiment, all the road sections on which the tested vehicle may travel obtained in the above steps are rendered to obtain road section polygons.

[0100] Understandably, during the rendering process, due to sampling, some lanes may overlap, and all lane sections are first unioned before rendering.

[0101] The road section polygon is divided into grids, all grid center points are taken as candidate points, and dense candidate points of possible road sections are output to reflect the possibility of the driving trajectory of the tested vehicle.

[0102] Take a section of the output road rectangle as an example, divide the road into grids with appropriate density (too high density increases the calculation pressure, too low density affects the prediction accuracy), and set the center point of each grid as the candidate point. The purpose of this step is to obtain appropriate candidate points. The target-driven trajectory prediction is to select from these candidate points and then determine the target point through subsequent processing.

[0103] like Figure 6 As shown, the present application also provides a target trajectory prediction device 200, which includes at least one software function module that can be stored in a storage module in the form of software or firmware or solidified in the operating system (OS) of the target trajectory prediction device 10. The processor 13 is used to execute the executable module stored in the storage module, such as the software function module and computer program included in the target trajectory prediction device 200.

[0104] The target trajectory prediction device 200 includes an information collection unit 210 and an analysis and processing unit 220. The functions of each unit may be as follows:

[0105] Information collection unit 210: used to collect vehicle status information of the tested vehicle during its travel and centerline position information of the road section, and perform coordinate system conversion and unification;

[0106] Analysis and processing unit 220: It is used to analyze and process the collected information, execute the road segment breadth-first search algorithm, and determine the road segment polygon of the possible driving lanes of the vehicle under test.

[0107] In this embodiment, the storage module can be, but is not limited to, random access memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, etc. In this embodiment, the storage module can be used to store the driving speed V, acceleration A, preset prediction time T, central position P of the vehicle under test, queue Q, list L, etc. collected by the vehicle under test. Of course, the storage module can also be used to store programs, and the processing module executes the program after receiving the execution instruction.

[0108] It can be understood that Figure 1 the structure of the target trajectory prediction device 10 shown in Figure 1 is only a schematic structural diagram, and the target trajectory prediction device 10 may also include more Figure 1 components than those shown.

[0109] It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described target trajectory prediction device 10 and target trajectory prediction device 200 can refer to the corresponding processes of each step in the foregoing method, and will not be elaborated here.

[0110] This application embodiment also provides a vehicle. The vehicle includes a vehicle body and the target trajectory prediction device 10 described in the above embodiment. The target trajectory prediction device 10 is deployed on the vehicle body. The target trajectory prediction device 10 can be used to implement the above target trajectory prediction method, which can improve the reliability of the trajectory during the operation of the vehicle, thereby contributing to improving driving safety.

[0111] This application embodiment also provides a computer-readable storage medium. A computer program is stored in the computer-readable storage medium. When the computer program runs on a computer, it causes the computer to execute the target trajectory prediction method described in the above embodiment.

[0112] Through the description of the above embodiments, those skilled in the art can clearly understand that this application can be implemented through hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of this application can be embodied in the form of a software product, and this software product can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a prediction device, or a network device, etc.) to execute the methods described in various implementation scenarios of this application.

[0113] In summary, the embodiments of this application provide a target trajectory prediction method, device, braking device, vehicle, and storage medium. In this solution, the processor 13 performs weighted calculations based on the collected driving speed V, acceleration A of the vehicle to be measured, and the preset prediction time T, obtains the dynamically searched range in real time according to the vehicle driving state, combines the preset road segment breadth-first search algorithm, obtains the lane segments that the vehicle to be measured may travel, and then performs rendering and divides the grid to output the dense candidate points of the possible segments. By monitoring the driving state information of the vehicle to be measured and the road, the search range is adjusted in real time, reducing the generation of a large amount of redundant data and saving computing power. Since the search range is adjusted in real time according to the driving state information of the vehicle to be measured, the accuracy of trajectory prediction is improved.

[0114] In the embodiments provided in this application, it should be understood that the disclosed devices, systems, and methods can also be implemented in other ways. The device, system, and method embodiments described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions, and operations of systems, methods, and computer program products according to multiple embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code, and the module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, the functional modules in each embodiment of this application can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0115] The above description is only for the embodiments of this application and is not intended to limit the protection scope of this application. For those skilled in the art, this application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the protection scope of this application.

Claims

1. A target trajectory prediction method, characterized in that: Applied to a target trajectory prediction device, the device includes a positioning module, a signal acquisition module and a processor. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The method includes: The positioning module establishes coordinates with the ground as the reference system. The signal acquisition module acquires the state information of the vehicle under test and sends the state information of the vehicle under test to the processor; The processor calculates the state information of the vehicle under test to determine the central position P of the vehicle under test and determines the road section where the vehicle under test is currently located; According to a preset search radius algorithm, combined with a preset prediction time T and the state information of the vehicle under test, determine the dynamic search range; According to the road section information where the vehicle under test is currently located and the dynamic search range, through a preset road section breadth-first search algorithm, predict and divide the road sections that the vehicle under test may travel; Render the predicted and divided road sections to obtain a road section polygon of the road sections that the vehicle under test may travel; Perform grid division on the road section polygon, take the center points of all grids as candidate points, and output dense candidate points of possible road sections; The signal acquisition module acquires the driving speed V, the heading H of the vehicle under test, and the acceleration A of the vehicle under test; According to a preset search radius algorithm, combined with a preset prediction time T and the state information of the vehicle under test, determine the dynamic search range, including: Perform weighted calculation based on the driving speed V, acceleration A, and a preset prediction time T to obtain a calculated search radius R dy ; A minimum search radius R0 is preset in the processor, and the calculated search radius R dy is compared with the minimum search radius R0; Take the maximum value as the actual search radius R, generate a semi-circular dynamic search range with the center position P as the center and the actual search radius R as the radius, and the heading H coincides with the angular bisector of the dynamic search range; According to the road section information where the vehicle under test is currently located and the dynamic search range, through a preset road section breadth-first search algorithm, predict and divide the road sections that the vehicle under test may travel, including; During driving, the processor sets different road section ids for the road section where the vehicle under test is currently located and the road sections involved in the dynamic search range; A queue Q and a list L are set in the processor. The queue Q is used to store the road section id where the vehicle under test is currently located and the ids of the adjacent road sections of the road section where the vehicle under test is currently located. The list L is used to store the ids of the possible road sections that have been searched; The road section id where the vehicle under test is currently located is added to the queue Q and the list L as the starting road section id respectively; During the forward movement of the vehicle under test, the processor sequentially adds the id of the road section where the vehicle under test is currently driving and the id of the critical road section to the queue Q; Loop through the queue Q, extract the road section id at the head of the queue Q, calculate the Euclidean distance from the center line of the road section at the head of the queue to the center position P, and compare it with the actual search radius R; If the Euclidean distance from the center line of the road section at the head of the queue to the center position P exceeds the actual search radius R, pop the road section id at the head of the queue from the queue Q; If the Euclidean distance from the center line of the lane segment at the head of the queue to the center position P does not exceed the actual search radius R, add the lane segment id at the head of the queue to the list L, and then pop the lane segment id at the head of the queue from the queue Q; When the elements in the queue Q are empty, output a list L containing all possible lane segments that the measured vehicle can travel.

2. The target trajectory prediction method according to claim 1, wherein: The signal acquisition module acquires the state information of the measured vehicle, including: The signal acquisition module acquires the body length positioning coordinates and the body width coordinates.

3. The target trajectory prediction method according to claim 2, wherein: The processor calculates the state information of the measured vehicle to determine the center position P of the measured vehicle and determines the lane segment where the measured vehicle is currently located, including: Construct a rectangle diagonal according to the body length positioning coordinates and the body width coordinates, and the coordinates of the intersection point of the rectangle diagonal are the center position P of the measured vehicle; The signal acquisition module acquires the coordinate arrays of the center lines of multiple lane segments near the measured vehicle, calculates the Euclidean distance from each center line of the lane segment to the center position P, and selects the lane segment where the center line of the lane segment closest to the center position P is located as the lane segment where the measured vehicle is located.

4. A target trajectory prediction method according to claim 1, characterized in that: Render the predicted lane segments to obtain the segment polygon of the lane segments that the measured vehicle may travel, including: Render the lane segments corresponding to all lane segment ids in the list L and output the segment polygon of the lane segments that the measured vehicle may travel.

5. A target trajectory prediction device, characterized in that: Applied to a target trajectory prediction device, the device includes a positioning module, a signal acquisition module, and a processor. The processor is used to process the data acquired by the signal acquisition module to improve the trajectory prediction accuracy. The device includes: An information collection unit: used to collect the vehicle state information and the center line position information of the lane segment during the travel of the measured vehicle, and perform coordinate system conversion and unification; An analysis and processing unit: used to analyze and process the collected information, execute the breadth-first search algorithm for road segments, and determine the segment polygon of the lane segments that the measured vehicle may travel; The signal acquisition module acquires the driving speed V, the heading H of the measured vehicle, and the acceleration A of the measured vehicle; The analysis and processing unit is used to determine the dynamic search range according to the preset search radius algorithm, in combination with the preset prediction time T and the state information of the measured vehicle, including: Perform weighted calculation based on the driving speed V, acceleration A, and a preset prediction time T to obtain a calculated search radius R dy ; A minimum search radius R0 is preset in the processor, and the calculated search radius R dy is compared with the minimum search radius R0; Take the maximum value as the actual search radius R, generate a semi-circular dynamic search range with the center position P as the center and the actual search radius R as the radius, and the heading H coincides with the angle bisector of the dynamic search range; The analysis and processing unit is also used to predict and divide the lane segments that the measured vehicle may travel through the preset breadth-first search algorithm for road segments according to the lane segment information where the measured vehicle is currently located and the dynamic search range, including; During the driving process, the processor sets different lane segment ids for the lane segment where the measured vehicle is currently located and the lane segments involved in the dynamic search range; A queue Q and a list L are set in the processor. The queue Q is used to store the lane segment id where the vehicle under test is currently located and the ids of the adjacent segments of the lane segment where the vehicle under test is currently located. The list L is used to store the lane segment ids of the possible travel routes that have been searched. The lane segment id where the vehicle under test is currently located is added to the queue Q and the list L respectively as the starting segment id. During the forward movement of the vehicle under test, the processor sequentially adds the lane segment id of the road section where the vehicle under test is traveling and the id of the critical section to the queue Q. Traverse the queue Q cyclically, extract the lane segment id at the head of the queue Q, calculate the Euclidean distance from the center line of the lane segment at the head of the queue to the central position P, and compare it with the actual search radius R. If the Euclidean distance from the center line of the lane segment at the head of the queue to the central position P exceeds the actual search radius R, the lane segment id at the head of the queue is popped out of the queue Q. If the Euclidean distance from the center line of the lane segment at the head of the queue to the central position P does not exceed the actual search radius R, the lane segment id at the head of the queue is added to the list L, and then the lane segment id at the head of the queue is popped out of the queue Q. When the elements in the queue Q are empty, a list L containing all the possible lane segments that the vehicle under test can travel is output.

6. A target trajectory prediction device, characterized in that: It includes a positioning module, a signal acquisition module, a processor, and a storage module. The processor is used to process the data collected by the signal acquisition module to improve the trajectory prediction accuracy. The storage module stores a computer program. When the computer program is executed by the processor, the target trajectory prediction device executes the method described in any one of claims 1-4.

7. A vehicle, characterized in that, The vehicle includes a vehicle body and the target trajectory prediction device described in claim 6. The target trajectory prediction device is arranged on the vehicle.

8. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium. When the computer program runs on a computer, the computer executes the method described in any one of claims 1-4.

Images