Track planning method, device and equipment and storage medium

Abstract

The invention discloses a trajectory planning method, apparatus and device, and a storage medium. The method comprises the steps of obtaining vehicle driving data and driving environment data of a target vehicle; inputting the vehicle driving data and the driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle; determining a current driving scene of the target vehicle based on the driving environment data; based on the current driving scene, respectively performing curvature smoothing on the plurality of predicted driving paths to obtain a plurality of candidate driving paths; and performing trajectory synthesis on the predicted speed distribution data and the candidate driving path to obtain a target driving trajectory. According to the method, the predicted driving path is optimized, and the candidate driving paths are screened in combination with the predicted speed distribution data to obtain the target driving track, so that the accuracy and safety of determining the driving track of the target vehicle can be improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and in particular to a trajectory planning method, device, equipment and storage medium. Background Art

[0002] For trajectory planning of autonomous vehicles, related technologies mainly generate multiple candidate trajectories based on the vehicle's historical driving data and surrounding environment information. These candidate trajectories provide possible future driving paths for the vehicle and provide decision-making references for downstream planning modules.

[0003] However, the trajectory generated by the trajectory prediction model in different scenarios may have motion trajectory defects, resulting in the vehicle being unable to pass according to the expected trajectory, thus affecting the accuracy and safety of trajectory planning. Summary of the invention

[0004] In view of the above-mentioned defects of the related art, the technical problem to be solved by the present invention is how to improve the accuracy and safety of determining the target driving trajectory.

[0005] In order to solve at least one of the technical problems mentioned above, the present invention discloses a trajectory planning method, device, equipment and storage medium.

[0006] According to one aspect of the present disclosure, a trajectory planning method is provided, comprising:

[0007] Acquire vehicle driving data and driving environment data of the target vehicle;

[0008] Inputting the vehicle driving data and the driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle;

[0009] Based on the driving environment data, determining a current driving scene of the target vehicle;

[0010] Based on the current driving scene, curvature smoothing is performed on the multiple predicted driving paths to obtain multiple candidate driving paths;

[0011] The predicted speed distribution data and the candidate driving paths are subjected to trajectory synthesis to obtain a target driving trajectory.

[0012] In some possible embodiments, determining the current driving scene of the target vehicle based on the driving environment data includes:

[0013] Determine the map location information corresponding to the target vehicle based on the vehicle driving data and the driving environment data;

[0014] Determine a plurality of sampling points based on the map location information, and calculate an average sampling point curvature of the plurality of sampling points;

[0015] The current driving scene is determined based on the average sampling point curvature.

[0016] In some possible embodiments, based on the current driving scene, curvature smoothing is performed on the multiple predicted driving paths to obtain multiple candidate driving paths, including:

[0017] Determining a preset curvature smoothing condition matching the current driving scene;

[0018] Based on the preset curvature smoothing condition, curvature smoothing is performed on the multiple predicted driving paths to obtain the multiple candidate driving paths.

[0019] In some possible embodiments, based on the preset curvature smoothing condition, curvature smoothing is performed on the multiple predicted driving paths respectively to obtain the multiple candidate driving paths, including:

[0020] Determine a plurality of first control points and a plurality of second control points corresponding to the plurality of predicted driving paths respectively; the first control point is a starting point of each predicted driving path, and the second control point is an end point of each predicted driving path;

[0021] Calculating curvature based on the plurality of predicted driving paths to obtain a plurality of average curvatures; each of the predicted driving paths corresponds to an average curvature;

[0022] Based on each average curvature, determining a plurality of intermediate control points from the corresponding predicted driving path;

[0023] Based on the first smoothing condition, the multiple first control points, the multiple second control points, and the multiple intermediate control points, curvature smoothing is performed on the multiple predicted driving paths to obtain the multiple candidate driving paths.

[0024] In some possible embodiments, based on the preset curvature smoothing condition, curvature smoothing is performed on the multiple predicted driving paths respectively to obtain the multiple candidate driving paths, including:

[0025] Get the preset tracking distance;

[0026] Based on the second smoothing condition and the preset tracking distance, curvature smoothing is performed on the multiple predicted driving paths respectively to obtain the multiple candidate driving paths.

[0027] In some possible embodiments, performing trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory includes:

[0028] Performing data smoothing on the predicted speed distribution data to obtain target speed distribution data;

[0029] Trajectory synthesis is performed based on the target speed distribution data and the candidate driving paths to obtain the target driving trajectory.

[0030] In some possible embodiments, performing trajectory synthesis based on the target speed distribution data and the candidate driving paths to obtain the target driving trajectory includes:

[0031] Performing trajectory synthesis based on the target speed distribution data and the candidate driving paths to obtain a plurality of driving trajectories to be screened;

[0032] The plurality of driving trajectories to be screened are screened based on preset trajectory screening conditions to obtain the target driving trajectory; and the number of the target driving trajectories is multiple.

[0033] According to a second aspect of the present disclosure, a trajectory planning device is provided, the device comprising:

[0034] A data acquisition module, used to acquire vehicle driving data and driving environment data of a target vehicle;

[0035] A data prediction module, used to input the vehicle driving data and the driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle;

[0036] A scene determination module, used to determine the current driving scene of the target vehicle based on the driving environment data;

[0037] A curvature smoothing module, used to perform curvature smoothing on the multiple predicted driving paths based on the current driving scene to obtain multiple candidate driving paths;

[0038] The trajectory determination module is used to perform trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory.

[0039] There may be other ways of dividing modules. The present application does not limit the way of dividing modules of the device for implementing the above functions. For example, the above-mentioned data acquisition module, data prediction module, scene determination module, curvature smoothing and trajectory determination module can be integrated / implemented as one or more processing modules.

[0040] According to a third aspect of the present disclosure, an electronic device is provided, the electronic device comprising a processor. The processor can execute any of the above-described trajectory planning methods. Optionally, the electronic device may further comprise a memory, wherein at least one instruction and at least one program are stored in the memory, and the at least one instruction and the at least one program are loaded and executed by the processor to implement the above-described trajectory planning method.

[0041] According to a fourth aspect of the present disclosure, a computer storage medium is provided, wherein at least one instruction and at least one program are stored in the computer storage medium, and the at least one instruction and the at least one program are loaded and executed by a processor to implement the trajectory planning method as described above. For example, the medium is a computer-readable storage medium.

[0042] According to a fifth aspect of the present disclosure, a chip system is provided, comprising at least one processor and at least one interface circuit, wherein the at least one interface circuit is used to perform transceiver functions and send instructions to the at least one processor, and when the at least one processor executes the instructions, the at least one processor executes the trajectory planning method as described in the first aspect and any one of the designs thereof.

[0043] According to a sixth aspect of the present disclosure, a computer program product is provided, comprising a computer program / instruction, wherein the computer program / instruction implements the trajectory planning method as described above when executed by a processor.

[0044] According to the seventh aspect of the present disclosure, a trajectory planning method is provided, which can be executed by an electronic device, and the electronic device can be at a component level (such as a chip) or a whole machine level (such as a vehicle). The method may include: obtaining a candidate driving path, wherein the candidate driving path is obtained by curvature smoothing of a predicted driving path; and controlling driving execution based on the candidate driving path.

[0045] Optionally, based on the candidate driving path, controlling driving execution includes: performing trajectory synthesis on the predicted speed distribution data and the candidate driving path to obtain a target driving trajectory.

[0046] The implementation of the present invention has the following beneficial effects:

[0047] In the present invention, the vehicle driving data and driving environment data corresponding to the target vehicle are input into the trajectory prediction model to obtain multiple predicted driving paths and corresponding predicted speed distribution data, and the current driving scene of the target vehicle is determined according to the driving environment data, and the predicted driving path is curvature smoothed according to the current driving scene to obtain a candidate driving path, so that the candidate driving path can meet the actual driving needs under the scenes corresponding to different path curvatures, avoiding the steering problem caused by insufficient trajectory curvature or sudden change of curvature, thereby improving the safety and accuracy of the candidate driving path. In addition, the candidate driving path is combined with the predicted speed distribution data, so that the obtained target driving trajectory can take into account the constraints of road conditions, traffic rules and dynamic environment (such as the speed of the preceding vehicle, the time of the traffic light, etc.), thereby improving the dynamic adaptability of the trajectory planning; at the same time, the speed distribution data is combined with the candidate driving path, which can avoid unreasonable acceleration, deceleration or excessive lateral acceleration, and thus can better meet the vehicle dynamics model, avoid loss of control due to excessive speed, thereby improving the vehicle driving safety and the comfort of the user riding in the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solution of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 A schematic diagram of a flow chart corresponding to the trajectory planning method provided in an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of a corresponding process for determining a current driving scenario provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of a process flow corresponding to trajectory curvature smoothing provided by an embodiment of the present invention;

[0052] Figure 4 A schematic diagram of a corresponding process for determining a target trajectory provided by an embodiment of the present invention;

[0053] Figure 5 A schematic diagram of the structure corresponding to the trajectory planning device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of this specification to clearly and completely describe the technical solutions in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work can fall within the scope of protection of the present invention.

[0055] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, such as a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products, or devices.

[0056] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0057] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0058] The term "and / or" herein is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C.

[0059] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0060] Figure 1The following is a flow chart of the trajectory planning method provided by the embodiment of the present invention; the execution subject may be any terminal capable of executing the trajectory planning method, such as a controller, a server, etc., or may be a corresponding module capable of implementing or cooperating with the implementation of the terminal function, such as a chip. Figure 1 , a trajectory planning method, comprising:

[0061] Step S101: Acquire vehicle driving data and driving environment data of a target vehicle;

[0062] In a feasible embodiment, the vehicle driving data may include the current position information and current speed information of the target vehicle; the driving environment data may include the corresponding position information and speed information of other vehicles around the target vehicle on the road, map data information, and traffic light information, lane line information, dynamic obstacle information, and static obstacle information obtained through sensors.

[0063] Among them, the map data information may include lane line information, such as the left boundary line, the right boundary line and the center line, etc. The lane line information includes the point coordinates and attributes (color, virtuality, type, etc.) of these lines; in addition, the lane line information may also include the connection relationship between lanes (previous, next, left adjacent, right adjacent). The map data information may also include intersection information. For example, for an intersection that does not include lane lines, the intersection information may include intersection contour information, intersection type information, connection relationship with non-intersections, etc.; at the same time, the map data information may also include road surface marking information, such as turn arrows on lanes, etc. There is partial overlap between the map data information and the information obtained by the sensor, such as lane line information, and the information obtained by the two acquisition methods can mutually verify the accuracy of the driving environment data acquisition.

[0064] Step S102: inputting vehicle driving data and driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle;

[0065] In a specific embodiment, after obtaining the above information, it is input into a pre-trained trajectory prediction model, and the trajectory prediction model infers the input information to obtain N predicted driving paths and M predicted speed distribution data, each predicted driving path can correspond to K predicted speed distribution data, that is, M = N*K, where the values ​​of K and N can be set according to specific needs, for example, K can be 6, N can be 10, the number of predicted driving paths is 10, and 1 predicted driving path can correspond to 6 predicted speed distribution data. In addition, after inputting the vehicle driving data and driving environment data into the prediction trajectory model, the predicted trajectory corresponding to the dynamic obstacle information and the surrounding vehicle information can also be obtained.

[0066] For the predicted speed distribution data, its form can be the position on the predicted driving path at a given time, such as 0.1s at 1 meter from the starting point on the path, and 0.2s at 3 meters from the starting point. Each predicted driving path and the corresponding speed distribution can be synthesized into a trajectory. Among them, the difference between a path and a trajectory is that a path is just a pure point, and a trajectory contains time, that is, the time at which each point is located. Setting each predicted driving path to correspond to K predicted speed distribution data can provide more possibilities for subsequent trajectory selection, thereby finding a more suitable trajectory and improving the safety, accuracy and adaptability of trajectory planning to specific form requirements.

[0067] Reasoning is the forward reasoning of the trajectory prediction model. After receiving the encoded data input in the specified format, the processor automatically performs calculations and outputs the calculation results. Since the embodiment of the present invention is for post-processing of the model output results, the structure and training process of the model itself are not described in detail.

[0068] Step S103: determining the current driving scene of the target vehicle based on the driving environment data;

[0069] In a specific embodiment, the current driving scene is determined, where the driving scene may include a large curve, a U-turn, a left or right turn, etc. The purpose is to distinguish scenes with high curvature in the predicted driving path and optimize them to make up for the defects in the output path. Therefore, left and right turns can be classified into one category and distinguished by different directions.

[0070] Figure 2 A schematic diagram of a flow chart corresponding to a trajectory curvature smoothing provided by an embodiment of the present invention; Figure 2 As shown, step S103 may specifically include:

[0071] Step S201: determining map location information corresponding to a target vehicle based on vehicle driving data and driving environment data;

[0072] In a specific embodiment, the position of the target vehicle in the map may be determined first according to the vehicle position information and the map data information. The map position information may be combined with the map data information to indicate that the target vehicle is located in a lane or at an intersection.

[0073] Step S202: determining a plurality of sampling points based on the map location information, and calculating an average sampling point curvature of the plurality of sampling points;

[0074] In a specific embodiment, when the target vehicle is located on a lane, the current position of the target vehicle is taken as the starting point, and the vehicle moves backward along the lane for S meters, and a sampling point is determined at intervals of x meters. For example, the vehicle moves backward for 100 meters and a sampling point is determined at intervals of 1 meter, thereby obtaining multiple sampling points, and calculating the average sampling point curvature of these sampling points.

[0075] Step S203: Determine the current driving scene based on the average sampling point curvature.

[0076] In a specific embodiment, the average sampling point curvature is compared with a preset average sampling point threshold, and if the average sampling point curvature is greater than the average sampling point threshold, it is determined that the target vehicle is in a curved path and its current driving scene is a curve. The average sampling point threshold can be preset according to different driving requirements.

[0077] In another specific embodiment, when the target vehicle is at an intersection, the lane or virtual lane at the intersection closest to the current point is searched, and the angle corresponding to the direction of the points at the beginning and end of the lane is calculated. If it exceeds a preset angle threshold, it is considered to be a left or right turn or a U-turn. Among them, the angle threshold can be preset according to different driving requirements. In addition, since the curvature of the path corresponding to the U-turn is much greater than the curvature of the path corresponding to the left or right turn, it is possible to further determine whether the current driving scene is a left or right turn or a U-turn based on the curvature value; for left and right turns, it can be distinguished based on the positive and negative signs of the average curvature.

[0078] Step S104: Based on the current driving scene, curvature smoothing is performed on the multiple predicted driving paths to obtain multiple candidate driving paths;

[0079] In a specific embodiment, since the predicted driving path output by the trajectory prediction model may have insufficient curvature or sudden changes, the predicted driving path can be smoothed to solve the problem of insufficient curvature or sudden changes in curvature, and a candidate driving path corresponding to each predicted driving path can be obtained. Determining the curvature smoothing method of the predicted driving path according to the driving scenario can improve the adaptability of the curvature smoothing of the predicted driving path, thereby improving the reliability of the smoothing result, i.e., the candidate driving path.

[0080] Figure 3 A schematic diagram of a flow chart corresponding to a trajectory curvature smoothing provided by an embodiment of the present invention, such as Figure 3 As shown, step S104 may include:

[0081] Step S301: determining a preset curvature smoothing condition matching the current driving scene;

[0082] In a specific embodiment, the preset curvature smoothing condition may include a first smoothing condition, a second smoothing condition and a third smoothing condition; wherein the first smoothing condition may correspond to large curves and left and right turn scenarios, and a high-order Bezier curve may be used for curvature smoothing; the second smoothing condition may correspond to a U-turn scenario, and a pure tracking method may be used for curvature smoothing; the third smoothing condition may correspond to a general scenario, i.e., a scenario other than left and right turns, large curves, and U-turn scenarios, and a high-order Bezier curve may be used for curvature smoothing; and then based on the preset curvature smoothing condition, curvature smoothing is performed on multiple predicted driving paths respectively to obtain multiple candidate driving paths.

[0083] Specifically, based on a preset curvature smoothing condition, curvature smoothing is performed on a plurality of predicted driving paths respectively to obtain a plurality of candidate driving paths, which may include:

[0084] Step S302: determining a plurality of first control points and a plurality of second control points corresponding to a plurality of predicted driving paths; the first control point is a starting point of each predicted driving path, and the second control point is an end point of each predicted driving path;

[0085] In a specific embodiment, for large curves and left and right turning scenarios, high-order Bezier curves are used for curvature smoothing. First, the starting point of each predicted driving path is determined as the first control point, and the second control point is determined as the ending point. The first control point and the second control point are respectively the first and last points of the smoothed curve.

[0086] Step S303: Calculating curvature based on multiple predicted driving paths to obtain multiple average curvatures; each predicted driving path corresponds to an average curvature;

[0087] Step S304: Based on each average curvature, determine a plurality of intermediate control points from the corresponding predicted driving path;

[0088] In a specific embodiment, the intermediate control points with smooth curvature can be determined according to the average curvature of each predicted driving path. Specifically, for any predicted driving path, the curvature of each point can be calculated first, and then the average curvature of the predicted driving path can be calculated. For the predicted driving path, each point corresponding to a curvature greater than the average curvature is set as an intermediate control point, and for the points corresponding to a curvature less than the average curvature, an intermediate control point is set every 2 points. The same method is used to determine the intermediate control points of each predicted driving path.

[0089] The selection of intermediate control points is determined according to the curvature, so that the control points are dense where the curvature is large and sparse where the curvature is small, thereby achieving curvature smoothing for each predicted driving path and reducing the impact of insufficient curvature on the accuracy of the predicted driving path.

[0090] Step S305: Based on the first smoothing condition, the plurality of first control points, the plurality of second control points and the plurality of intermediate control points, curvature smoothing is performed on the plurality of predicted driving paths respectively to obtain a plurality of candidate driving paths.

[0091] In a specific embodiment, a high-order Bezier curve is used, with the first control point and the second control point of each predicted driving path as the first and last points of curvature smoothing, respectively, and multiple intermediate control points are used to perform curvature smoothing on each predicted driving path, and a candidate driving path corresponding to each predicted driving path is obtained.

[0092] Alternatively, based on a preset curvature smoothing condition, curvature smoothing is performed on multiple predicted driving paths respectively to obtain multiple candidate driving paths, including:

[0093] Step S306: Obtaining a preset tracking distance;

[0094] Step S307: Based on the second smoothing condition and the preset tracking distance, curvature smoothing is performed on the multiple predicted driving paths respectively to obtain multiple candidate driving paths.

[0095] In a specific embodiment, since the curvature of the U-turn path is larger than the curvature of the trajectory corresponding to other scenarios, the pure tracking method is selected to smooth the curvature of the predicted driving path. The pure tracking method can not only smooth the curvature, but also increase the curvature, thereby making up for the problem of insufficient curvature of the path corresponding to the U-turn, thereby improving the accuracy, reliability and safety of the driving trajectory determination.

[0096] Specifically, the preset tracking distance, i.e., the preview distance, can be set according to specific driving requirements. For example, the preset tracking distance can be set to 20m.

[0097] Alternatively, based on a preset curvature smoothing condition, curvature smoothing is performed on multiple predicted driving paths respectively to obtain multiple candidate driving paths, including:

[0098] Step S302: determining a plurality of first control points and a plurality of second control points corresponding to a plurality of predicted driving paths; the first control point is a starting point of each predicted driving path, and the second control point is an end point of each predicted driving path;

[0099] Step S308: determining a plurality of intermediate control points from the plurality of predicted driving paths at a preset control point spacing;

[0100] Step S309: Based on the third smoothing condition, the first control point, the second control point and the intermediate control point, curvature smoothing is performed on the plurality of predicted driving paths respectively to obtain a plurality of candidate driving paths.

[0101] In a specific embodiment, when the driving scene is a general scene, its curvature smoothness corresponds to the third smoothness condition, and the determination of the first control point and the second control point is the same as the determination method of the first control point and the second control point in the first smoothness condition, which will not be repeated here.

[0102] In general scenarios, multiple intermediate control points can be determined from each predicted driving path based on each point on each predicted driving path, that is, the preset control point spacing is the distance corresponding to one point, and then multiple intermediate control points corresponding to the multiple predicted driving paths are determined.

[0103] Step S105: performing trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory.

[0104] In a specific embodiment, the candidate driving paths and speed distribution data are used to synthesize the trajectory to obtain the target driving trajectory. The planned points in these driving paths are connected to obtain a broken line composed of many line segments. The sum of the distances of these broken lines is the distance to the starting point. Then, combined with the distance from the starting point given by the speed distribution, the coordinates can be calculated by interpolation of these broken lines. The coordinates are the trajectory points. Each corresponding trajectory point on a candidate driving path is connected together to obtain the target driving trajectory. For example, the distance between the third point and the first point (starting point) is 10 meters, and the distance between the fourth point and the starting point is 13 meters. The speed distribution gives 0.5s at 11 meters from the starting point. Then the coordinates of 11 meters can be calculated by interpolation using the third and fourth points. This coordinate is the coordinate of 0.5s, which is a trajectory point.

[0105] Figure 4 A schematic diagram of a target trajectory determination process provided by an embodiment of the present invention; Figure 4 As shown, step S105 may specifically include:

[0106] Step S401: smoothing the predicted speed distribution data to obtain target speed distribution data;

[0107] In a specific embodiment, before performing trajectory synthesis, the predicted speed distribution data may be smoothed. One-dimensional and polynomial smoothing may be used to smooth the predicted speed distribution data, respectively, so as to make the speed change more continuous, reduce the trajectory acceleration fluctuation caused by the speed mutation, and reduce data noise.

[0108] Specifically, for all predicted speed distribution data, the Savizkg-Gollag filter algorithm can be used to smooth the target speed distribution data; further, the acceleration and jerk of the target speed distribution data are calculated to determine whether there is a discontinuous change exceeding the corresponding threshold; if there is, it means that the target speed distribution data after data smoothing does not meet the requirements, and the target speed distribution data that does not meet the requirements is smoothed again to reduce the overall noise. Among them, the thresholds corresponding to the acceleration and jerk can be percentages, for example, the acceleration difference between every two points is calculated, all the differences are averaged, and then these differences are checked to see if there are any differences that exceed the average by 2 or 3 times. If so, it is considered that the discontinuity is relatively high.

[0109] Further, performing trajectory synthesis based on the target speed distribution data and the candidate driving paths to obtain the target driving trajectory may include:

[0110] Step S402: synthesizing trajectories based on the target speed distribution data and the candidate driving paths to obtain a plurality of driving trajectories to be screened;

[0111] In a specific embodiment, trajectory synthesis can use target speed distribution data and candidate driving paths, thereby reducing the noise impact of predicted speed distribution data and the noise impact of predicted driving paths, and the obtained multiple driving trajectories to be screened are more reliable and safer than unprocessed data.

[0112] Step S403: screening multiple driving trajectories to be screened based on preset trajectory screening conditions to obtain a target driving trajectory; the number of target driving trajectories is multiple.

[0113] In a specific embodiment, multiple driving trajectories to be screened can be sorted according to preset trajectory screening conditions, and multiple target driving trajectories that meet the conditions can be screened. Specifically, the preset trajectory screening conditions may include obstacle distance, vehicle driving speed, and traffic rules. For example, the closer the distance to the obstacle, the lower the score; the slower the speed, the lower the score; the more traffic rules are violated (such as red lights, solid lines, etc.), the lower the score, etc. After sorting the multiple driving trajectories to be screened according to the scores, the first Q can be selected as the target driving trajectories, for example, Q can be 3. Selecting multiple driving trajectories to be screened as target driving trajectories can provide more selection space and serve as alternatives to each other to improve the reliability and safety of trajectory planning.

[0114] The embodiment of the present invention also provides a trajectory planning device, such as Figure 5 As shown, the device comprises:

[0115] The data acquisition module 510 is used to acquire the vehicle driving data and driving environment data of the target vehicle;

[0116] The data prediction module 520 is used to input the vehicle driving data and driving environment data into the trajectory prediction model to obtain multiple predicted driving paths and multiple predicted speed distribution data corresponding to the target vehicle;

[0117] A scene determination module 530, for determining a current driving scene of a target vehicle based on driving environment data;

[0118] The curvature smoothing module 540 is used to perform curvature smoothing on multiple predicted driving paths based on the current driving scene to obtain multiple candidate driving paths;

[0119] The trajectory determination module 550 is used to perform trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory.

[0120] In some other embodiments, the scene determination module 530 further includes:

[0121] A location determination module, used to determine the map location information corresponding to the target vehicle based on the vehicle driving data and driving environment data;

[0122] A data calculation module, used to determine multiple sampling points based on map location information, and calculate the average sampling point curvature of the multiple sampling points;

[0123] The information determination module is used to determine the current driving scene based on the average sampling point curvature.

[0124] In some other embodiments, the curvature smoothing module 540 further includes:

[0125] A first condition determination module, used to determine a preset curvature smoothing condition matching the current driving scene;

[0126] The path smoothing module is used to perform curvature smoothing on multiple predicted driving paths based on preset curvature smoothing conditions to obtain multiple candidate driving paths.

[0127] In some other embodiments, the path smoothing module further includes:

[0128] A first control point determination module, used to determine a plurality of first control points and a plurality of second control points corresponding to a plurality of predicted driving paths; the first control point is a starting point of each predicted driving path, and the second control point is an end point of each predicted driving path;

[0129] A curvature calculation module, used to calculate curvature based on multiple predicted driving paths to obtain multiple average curvatures; each predicted driving path corresponds to an average curvature;

[0130] A second control point determination module, for determining a plurality of intermediate control points from a corresponding predicted driving path based on each average curvature;

[0131] The first path processing module is used to perform curvature smoothing on multiple predicted driving paths based on a first smoothing condition, multiple first control points, multiple second control points, and multiple intermediate control points to obtain multiple candidate driving paths.

[0132] In some other embodiments, the path smoothing module further includes:

[0133] A distance acquisition module is used to obtain a preset tracking distance;

[0134] The second path processing module is used to perform curvature smoothing on the multiple predicted driving paths based on the second smoothing condition and the preset tracking distance to obtain multiple candidate driving paths.

[0135] In some other embodiments, the trajectory determination module 550 further includes:

[0136] A data smoothing module is used to smooth the predicted speed distribution data to obtain the target speed distribution data;

[0137] The first trajectory synthesis module is used to perform trajectory synthesis based on the target speed distribution data and the candidate driving paths to obtain a target driving trajectory.

[0138] In some other embodiments, the first trajectory synthesis module further includes:

[0139] A second trajectory synthesis module is used to synthesize trajectories based on the target speed distribution data and the candidate driving paths to obtain multiple driving trajectories to be screened;

[0140] The trajectory screening module is used to screen multiple driving trajectories to be screened based on preset trajectory screening conditions to obtain a target driving trajectory; the number of target driving trajectories is multiple.

[0141] The device and method embodiments in the device embodiment are based on the same inventive concept and are used to implement the above-mentioned trajectory planning method.

[0142] An embodiment of the present invention also provides an electronic device, comprising: a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement any trajectory planning method as described in the method embodiments.

[0143] An embodiment of the present invention also provides a storage medium, which can be set in a server to store at least one instruction, at least one program, code set or instruction set for implementing the trajectory planning method in the method embodiment, and at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement any trajectory planning method as described in the method embodiment.

[0144] Optionally, in an embodiment of the present invention, the above-mentioned storage medium may be located in at least one of a plurality of network servers of a computer network. Optionally, in an embodiment of the present invention, the above-mentioned storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0145] An embodiment of the present invention further provides a computer program product, including a computer program and / or instructions, which implements any trajectory planning method in the method embodiment when the computer program and / or instructions are executed by a processor.

[0146] As can be seen from the embodiments provided by the present invention, in the present invention, the vehicle driving data and driving environment data corresponding to the target vehicle are input into the trajectory prediction model to obtain multiple predicted driving paths and corresponding predicted speed distribution data, and the current driving scene of the target vehicle is determined according to the driving environment data, and the predicted driving path is curvature smoothed according to the current driving scene to obtain a candidate driving path, so that the candidate driving path can meet the actual driving needs under different scenes, avoiding the steering problem caused by insufficient trajectory curvature or sudden change of curvature, thereby improving the safety and accuracy of the candidate driving path. In addition, combining the candidate driving path with the predicted speed distribution data can accurately plan the driving state of the target vehicle, thereby further improving the safety and accuracy of the candidate driving path.

[0147] It should be noted that various embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes are obvious to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The choice of terms used in this article is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A trajectory planning method, characterized in that: The method comprises: Acquire vehicle driving data and driving environment data of the target vehicle; Inputting the vehicle driving data and the driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle; Based on the driving environment data, determining a current driving scene of the target vehicle; Based on the current driving scene, curvature smoothing is performed on the multiple predicted driving paths to obtain multiple candidate driving paths; The predicted speed distribution data and the candidate driving paths are subjected to trajectory synthesis to obtain a target driving trajectory.

2. A trajectory planning method according to claim 1, characterized in that: The determining, based on the driving environment data, a current driving scene of the target vehicle includes: Determine the map location information corresponding to the target vehicle based on the vehicle driving data and the driving environment data; Determine a plurality of sampling points based on the map location information, and calculate an average sampling point curvature of the plurality of sampling points; The current driving scene is determined based on the average sampling point curvature.

3. A trajectory planning method according to claim 2, characterized in that: Based on the current driving scene, curvature smoothing is performed on the multiple predicted driving paths to obtain multiple candidate driving paths, including: Determining a preset curvature smoothing condition matching the current driving scene; Based on the preset curvature smoothing condition, curvature smoothing is performed on the multiple predicted driving paths to obtain the multiple candidate driving paths.

4. A trajectory planning method according to claim 3, characterized in that: The preset curvature smoothing condition includes a first smoothing condition; The step of performing curvature smoothing on the plurality of predicted driving paths based on the preset curvature smoothing condition to obtain the plurality of candidate driving paths comprises: Determine a plurality of first control points and a plurality of second control points corresponding to the plurality of predicted driving paths respectively; the first control point is a starting point of each predicted driving path, and the second control point is an end point of each predicted driving path; Calculating curvature based on the plurality of predicted driving paths to obtain a plurality of average curvatures; each of the predicted driving paths corresponds to an average curvature; Based on each average curvature, determining a plurality of intermediate control points from the corresponding predicted driving path; Based on the first smoothing condition, the multiple first control points, the multiple second control points, and the multiple intermediate control points, curvature smoothing is performed on the multiple predicted driving paths to obtain the multiple candidate driving paths.

5. A trajectory planning method according to claim 3, characterized in that: The preset curvature smoothing condition includes a second smoothing condition; The step of performing curvature smoothing on the plurality of predicted driving paths based on the preset curvature smoothing condition to obtain the plurality of candidate driving paths comprises: Get the preset tracking distance; Based on the second smoothing condition and the preset tracking distance, curvature smoothing is performed on the multiple predicted driving paths respectively to obtain the multiple candidate driving paths.

6. A trajectory planning method according to claim 1, characterized in that: The step of performing trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory includes: Performing data smoothing on the predicted speed distribution data to obtain target speed distribution data; Trajectory synthesis is performed based on the target speed distribution data and the candidate driving paths to obtain the target driving trajectory.

7. A trajectory planning method according to claim 6, characterized in that: The step of synthesizing a trajectory based on the target speed distribution data and the candidate driving paths to obtain the target driving trajectory includes: Performing trajectory synthesis based on the target speed distribution data and the candidate driving paths to obtain a plurality of driving trajectories to be screened; The plurality of driving trajectories to be screened are screened based on preset trajectory screening conditions to obtain the target driving trajectory; and the number of the target driving trajectories is multiple.

8. A trajectory planning device, characterized in that: The device comprises: A data acquisition module, used to acquire vehicle driving data and driving environment data of a target vehicle; A data prediction module, used to input the vehicle driving data and the driving environment data into a trajectory prediction model to obtain a plurality of predicted driving paths and a plurality of predicted speed distribution data corresponding to the target vehicle; A scene determination module, used to determine the current driving scene of the target vehicle based on the driving environment data; A curvature smoothing module, used to perform curvature smoothing on the multiple predicted driving paths based on the current driving scene to obtain multiple candidate driving paths; The trajectory determination module is used to perform trajectory synthesis on the predicted speed distribution data and the candidate driving paths to obtain a target driving trajectory.

9. An electronic device, characterized in that: The electronic device includes a processor and a memory, wherein the memory stores at least one instruction and at least one program, and the at least one instruction and the at least one program are loaded and executed by the processor to implement the trajectory planning method according to any one of claims 1 to 7.

10. A computer storage medium, characterized in that: The computer storage medium stores at least one instruction and at least one program, and the at least one instruction and the at least one program are loaded and executed by the processor to implement the trajectory planning method according to any one of claims 1 to 7.

11. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the trajectory planning method according to any one of claims 1 to 7 is implemented.

Images