A parking trajectory selection method and device, a vehicle and a storage medium

Abstract

The application discloses a parking trajectory selection method and device, a vehicle and a storage medium. The method comprises the following steps: demarcating an evaluation area of a parking range according to a terminal point, and acquiring different parking points and corresponding initial values in the evaluation area; constructing a retriever in the time and position dimensions according to the predicted trajectory of an obstacle in a vehicle driving environment; taking the lateral and longitudinal speed change rate of the vehicle as a constraint of dynamic programming, taking the lateral and longitudinal acceleration change rate of the vehicle and the distance from the obstacle as a transition cost, and taking different parking points and corresponding initial values as a terminal point to perform multi-terminal dynamic programming, so as to obtain the numerical value of each point in a T-V-S-L four-dimensional graph; taking the numerical value of the point closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as a trajectory evaluation value, and selecting the lowest value as an actual parking trajectory. In this way, the distance between the parking point and the terminal point and the safety and comfort degree are taken as consideration factors to comprehensively evaluate different parking trajectories, which is helpful for automatically evaluating the parking trajectory.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of vehicles, in particular to a parking trajectory selection method and device, a vehicle and a storage medium. BACKGROUND

[0002] In recent years, automatic driving technology has developed rapidly, and its goal is usually to control the vehicle to autonomously travel along the road, to reach the destination as soon as possible while ensuring the safety of the vehicle, and to ensure that it does not directly or indirectly threaten the safety of other road users.

[0003] To achieve the above goal, the automatic driving software needs multiple key systems, and the trajectory planning system is one of them. The goal of trajectory planning is to plan a trajectory that meets the requirements of vehicle dynamics. This trajectory needs to be able to avoid surrounding obstacles (vehicles, pedestrians, static obstacles, etc.), and to meet the instructions of the decision layer (maintain the lane, change lanes, and park on the side of the road). Stable and reliable trajectory planning is one of the foundations to ensure the availability of automatic driving functions. Most solutions will provide multiple different trajectories, and then select the most reasonable one as the actual execution trajectory.

[0004] However, in complex scenarios, there is a problem that different parking trajectories cannot be evaluated, and it is difficult to select the optimal parking trajectory.

[0005] Therefore, how to evaluate different parking trajectories in complex scenarios while considering the distance from the end point, the safety level, and the comfort level to automatically select the optimal parking trajectory has become a problem to be solved. SUMMARY

[0006] The main purpose of the present application is to provide a parking trajectory selection method and device, a vehicle and a storage medium, which can evaluate different parking trajectories in complex scenarios while considering the distance from the end point, the safety level, and the comfort level to automatically select the optimal parking trajectory.

[0007] To achieve the above purpose, the present application provides a parking trajectory selection method, which comprises the following steps:

[0008] S1: According to the position of the end point, the evaluation area of the parking range is drawn, and the different parking point positions and their corresponding initial values in the evaluation area are obtained;

[0009] S2: According to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment, a retriever in the time and position dimensions is constructed, and the distance between the vehicle and the obstacle is retrieved;

[0010] S3: taking the lateral velocity change rate and the longitudinal velocity change rate of the vehicle as constraints of dynamic programming, taking the lateral acceleration change rate, the longitudinal acceleration change rate and the distance to the obstacle as transition costs, taking the different parking point positions and the corresponding initial values as the end points of dynamic programming, performing multi-endpoint dynamic programming, constructing a T-V-S-L four-dimensional graph and obtaining the values of each point in the T-V-S-L four-dimensional graph;

[0011] S4: taking the values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as trajectory evaluation values, traversing the evaluation values, and confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0012] Optionally, the step S1 comprises:

[0013] obtaining an end point position;

[0014] obtaining a parking point position at the end point position; wherein the parking point position comprises a determined parking point position and / or a possible parking point position;

[0015] delimiting a rectangular region as an evaluation region of a parking range according to the parking point position.

[0016] Optionally, the step S1 further comprises:

[0017] obtaining the Manhattan distance between the end point position and different parking point positions in the evaluation region;

[0018] taking the square value of each Manhattan distance as the initial value corresponding to the different parking point positions.

[0019] Optionally, the step S2 comprises:

[0020] obtaining the predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment;

[0021] interpolating the predicted trajectory of the obstacle to obtain a sampling result under discrete time sampling; wherein the sampling result comprises the position and the speed of each obstacle;

[0022] constructing a retriever in the time and position dimensions according to the sampling result.

[0023] Optionally, the step S2 further comprises:

[0024] obtaining a first position of the vehicle at a first time;

[0025] obtaining the distance between the vehicle and the obstacle at the first position according to the retriever.

[0026] Optionally, the step S3 comprises:

[0027] C1: taking the discrete time as the step of dynamic programming, and performing multi-endpoint dynamic programming according to time from large to small;

[0028] C2: taking the lateral speed change rate and the longitudinal speed change rate of the vehicle as the constraint of dynamic programming, and taking the lateral acceleration change rate, the longitudinal acceleration change rate and the distance from the vehicle to the obstacle as the transition cost;

[0029] C3: taking the different parking point positions and the corresponding initial values as the dynamic programming endpoints, constructing a T-V-S-L four-dimensional graph and obtaining the value of each point in the T-V-S-L four-dimensional graph.

[0030] Optionally, the step S4 comprises:

[0031] obtaining the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph;

[0032] taking the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as the trajectory evaluation value;

[0033] traversing the evaluation value to obtain the minimum evaluation value;

[0034] confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0035] In addition, to achieve the above-mentioned purpose, the application further provides a parking trajectory selection device, which comprises:

[0036] a division unit configured to divide an evaluation area of a parking range according to endpoint positions, and obtain different parking point positions and corresponding initial values in the evaluation area;

[0037] a first obtaining unit configured to construct a retriever in the time and position dimensions according to the predicted trajectories of static obstacles and / or dynamic obstacles in a vehicle driving environment, and retrieve the distance from the vehicle to the obstacles;

[0038] a second obtaining unit configured to take the lateral speed change rate and the longitudinal speed change rate of the vehicle as the constraint of dynamic programming, take the lateral acceleration change rate, the longitudinal acceleration change rate and the distance from the vehicle to the obstacles as the transition cost, perform multi-endpoint dynamic programming by taking the different parking point positions and the corresponding first reward values as the dynamic programming endpoints, construct a T-V-S-L four-dimensional graph and obtain the value of each point in the T-V-S-L four-dimensional graph;

[0039] a trajectory confirming unit configured to take the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as the trajectory evaluation value, traverse the evaluation value, and confirm the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0040] In addition, to achieve the above object, the application further provides a vehicle, comprising a memory, a processor and a parking trajectory selection program stored in the memory and executable on the processor, the parking trajectory selection program being configured to implement the steps of the parking trajectory selection method as described above.

[0041] In addition, to achieve the above object, the application further provides a computer readable storage medium, having a computer program stored thereon, the computer program being executable by a processor to implement the steps of the parking trajectory selection method as described above.

[0042] According to the technical scheme of the application, the evaluation area of the parking range is determined according to the end position, different parking point positions and corresponding initial values in the evaluation area are obtained, a retriever in the time and position dimensions is constructed according to the predicted trajectories of the static obstacles and / or dynamic obstacles in the vehicle driving environment, the distance between the vehicle and the obstacles is retrieved, the lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as the constraints of dynamic programming, the lateral acceleration change rate, the longitudinal acceleration change rate and the distance to the obstacles of the vehicle are taken as the transition cost, the different parking point positions and the corresponding initial values are taken as the end points of dynamic programming, a T-V-S-L four-dimensional graph is constructed and the values of each point in the T-V-S-L four-dimensional graph are obtained, the values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph are taken as the trajectory evaluation values, the evaluation values are traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory. In this way, the comprehensive evaluation method of different trajectories facing different end points is realized by the multi-endpoint dynamic programming algorithm, the distance between the parking point and the end point, the safety degree and the comfort degree are taken as the consideration factors to comprehensively evaluate different parking trajectories, and the evaluation of the parking trajectory is automatically realized. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 A flowchart of a parking trajectory selection method provided by the application.

[0044] Figure 2 A flowchart of step S1 of a parking trajectory selection method provided by the application.

[0045] Figure 3 A flowchart of step S2 of a parking trajectory selection method provided by the application.

[0046] Figure 4 A flowchart of step S3 of a parking trajectory selection method provided by the application.

[0047] Figure 5A flowchart of step S4 of the parking trajectory selection method provided by the present application is shown in the figure.

[0048] Figure 6 A structure block diagram of the parking trajectory selection device embodiment of the present application is shown in the figure.

[0049] Figure 7 A vehicle structure diagram of the hardware running environment involved in the embodiment scheme of the present application is shown in the figure.

[0050] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0051] In order to make the technical problems, technical solutions and beneficial effects of the present application more clear, specific embodiments of the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0052] In the subsequent description, the suffix such as "module", "component" or "unit" used to represent elements is only for the convenience of the description of the present application, and has no specific meaning. Therefore, "module", "component" or "unit" can be used mixedly.

[0053] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily describe a specific order or sequence.

[0054] Please refer to Figure 1 The present application provides a parking trajectory selection method, which comprises the following steps:

[0055] Step S1: According to the end position, the evaluation area of the parking range is drawn, and the different parking point positions and their corresponding initial values in the evaluation area are obtained.

[0056] Step S2: According to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment, a retriever in time and position dimension is constructed, and the distance between the vehicle and the obstacle is retrieved.

[0057] Step S3: The lateral velocity change rate and longitudinal velocity change rate of the vehicle are taken as the constraints of dynamic programming, the lateral acceleration change rate, longitudinal acceleration change rate and distance to the obstacle of the vehicle are taken as the transition cost, and the different parking point positions and their corresponding initial values are taken as the dynamic programming end point for multi-endpoint dynamic programming, to construct a T-V-S-L four-dimensional graph and obtain the numerical value of each point in the T-V-S-L four-dimensional graph.

[0058] Step S4: taking the value of the point closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as an evaluation value, traversing the evaluation value, and confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0059] In the embodiment of the present application, an evaluation area of a parking range is demarcated according to a terminal position, different parking point positions and corresponding initial reward values in the evaluation area are obtained, a retriever in time and position dimensions is constructed according to a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment, a vehicle-obstacle distance is retrieved, a lateral velocity change rate and a longitudinal velocity change rate of the vehicle are taken as constraints of dynamic programming, a lateral acceleration change rate, a longitudinal acceleration change rate and the vehicle-obstacle distance are taken as transition costs, the different parking point positions and the corresponding first reward values are taken as multiple-terminal dynamic programming endpoints, and a T-V-S-L four-dimensional graph is constructed. The value of each point in the T-V-S-L four-dimensional graph is the sum of the initial value and the transition cost. The value of the point closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph is taken as a trajectory evaluation value, the evaluation value is traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory. In this way, the present application realizes a comprehensive evaluation method for different trajectories facing different terminal points by using a multiple-terminal dynamic programming algorithm, and takes the distance between the parking point and the terminal point, the safety degree and the comfort degree as consideration factors to comprehensively evaluate different parking trajectories, which helps to automatically realize the evaluation of the parking trajectory.

[0060] As a specific embodiment, the step S1 can further refer to 2, as shown in the figure, the step S1 specifically includes: Figure 2

[0061] Step 101: demarcating an evaluation area of a parking range according to a terminal position.

[0062] In the embodiment, the terminal position can be obtained first, and then a parking point position at the terminal position is obtained. The parking point position includes a determined parking point position and / or a possible parking point position. Finally, a rectangular area demarcated according to the parking point position is taken as an evaluation area of a parking range.

[0063] For example, the terminal position can be X mall, the determined parking point position can be a parking lot inside X mall, and the possible parking point position can be a public parking lot 200 meters away from X mall, that is, there is a terminal position A (x1, y1) which is X mall, there is a parking point position B1 (x2, y2) which is a parking lot inside X mall, and there is a parking point position B2 (x3, y3) which is a public parking lot 200 meters away from X mall. Then, the rectangular area composed of the parking point positions can be the area of X mall plus the area within a range of 200 meters, which is taken as C area, as an evaluation area of a parking range. ​

[0064] Step 102: Obtain different parking point positions in the evaluation area and their corresponding initial values.

[0065] In this embodiment, the Manhattan distance between the end point position and different parking point positions in the evaluation area can be obtained; and the square value of each Manhattan distance is taken as the initial value corresponding to the different parking point position.

[0066] It should be noted that the Manhattan distance is also called the city block distance or L1 distance, which is a measurement method for calculating the distance between two points in a plane. The Manhattan distance can only move along the horizontal and vertical directions, so the shortest path by moving along the grid lines is the Manhattan distance.

[0067] For example, for two points on a plane, the end point position A(x1, y1) and the parking point position B1(x2, y2), the Manhattan distance (denoted as d(A, B1)) between them can be calculated by the following formula:

[0068] d(A, B1) = |x1-x2| + |y1-y2|

[0069] The formula represents the sum of the absolute values of the coordinate differences of the two points in the X-axis and Y-axis directions.

[0070] For the end point position A(x1, y1) and the parking point position B2(x3, y3) on the plane, the Manhattan distance (denoted as d(A, B2)) between them can be calculated by the following formula:

[0071] d(A, B2) = |x1-x3| + |y1-y3|

[0072] After obtaining each Manhattan distance, the square value of each Manhattan distance is taken as the initial value corresponding to the different parking point position.

[0073] For example, the initial value corresponding to the parking point position B1(x2, y2) is the square of d(A, B1), that is, d(A, B1) 2 ; the initial value corresponding to the parking point position B2(x3, y3) is the square of d(A, B2), that is, d(A, B2) 2 .

[0074] Therefore, the relationship between the end point position, different parking point positions, corresponding Manhattan distances and initial values can be shown in the following table:

[0075]

[0076] As a specific embodiment, the step S2 can further refer to 3, as shown in Figure 3 The step S2 specifically includes:

[0077] Step 201: constructing a retriever in time and position dimensions according to the predicted trajectory of the static obstacle and / or the dynamic obstacle in the vehicle driving environment.

[0078] In this embodiment, first, the predicted trajectory of the static obstacle and / or the dynamic obstacle in the vehicle driving environment is acquired; then, the predicted trajectory of the obstacle in the vehicle driving environment is interpolated to obtain a sampling result under discrete time sampling; wherein, the sampling result includes the position and speed of each obstacle. Preferably, a retriever in time and position dimensions can be constructed according to the sampling result.

[0079] Step 202: retrieving the distance between the vehicle and the obstacle according to the retriever.

[0080] In this embodiment, the first position of the vehicle at the first time can be acquired; then, the distance between the vehicle and the obstacle at the first position can be acquired according to the retriever. Specifically, the sampling result can be retrieved to obtain the distance of the nearest obstacle at the straight-line distance of the position at the first time, such as time ti, and the first position, such as position sl.

[0081] It should be noted that since the frenet coordinate system is also called the road coordinate system, in this embodiment, the SL coordinate system can be constructed with a lane center line to obtain the sl coordinates of each point in the road plane, and the coordinate axes are perpendicular to each other and divided into the s direction (i.e. the normal direction of the reference line, referred to as the longitudinal direction) and the l axis direction (i.e. the tangent direction along the road reference line, referred to as the transverse direction).

[0082] As a specific embodiment, the step S3 can further refer to 4, as shown in Figure 4 The step S3 specifically includes:

[0083] Step C1: taking discrete time as the step length of dynamic programming, and performing multi-endpoint dynamic programming according to time from large to small.

[0084] Step C2: taking the transverse speed change rate and the longitudinal speed change rate of the vehicle as the constraints of dynamic programming; and taking the transverse acceleration change rate, the longitudinal acceleration change rate, and the distance to the obstacle as the transition cost.

[0085] Step C3: taking the different parking point positions and their corresponding initial values as the endpoints of dynamic programming, constructing a T-V-S-L four-dimensional graph, and obtaining the values of each point in the T-V-S-L four-dimensional graph.

[0086] In this embodiment, the different parking point positions in step C3 and their corresponding initial values are obtained in step S1, which will not be described herein. As known by those skilled in the art, the lateral speed change rate, the longitudinal speed change rate, the lateral acceleration change rate and the longitudinal acceleration change rate of the vehicle can be calculated after the lateral speed, the longitudinal speed, the lateral acceleration and the longitudinal acceleration are obtained, which will not be described herein. In some optional embodiments, zero speed can also be used as the dynamic programming end point, which will not be described herein.

[0087] In this embodiment, the T-V-S-L four-dimensional graph can also be written as (t, v, s, l). The construction of (t, v, s, l) is based on the lane center line in the map, and the sl coordinates of each point on the road plane can be obtained by constructing the sl coordinate system with a lane center line. t and v are two dimensions of the vehicle state space. t is from 0 s to a certain time in the future (for example, 60 s), and v can be in the range of -15 m / s to +15 m / s. Discrete sampling in each dimension can obtain discrete (t, v, s, l).

[0088] In this embodiment, the mapping relationship from (t, v, s, l) to the evaluation value is a four-dimensional graph. However, the content of the mapping relationship is calculated by dynamic programming.

[0089] For example, it is known that the parking point is located at (s0, l0), then it can be assumed that the initial value of (60, 0, s0, l0) is 0, and the value of (59, v0, s1, l0) at the last time is 0 + longitudinal displacement cost + distance from obstacle cost + longitudinal comfort cost + lateral comfort cost.

[0090] wherein the longitudinal displacement cost is the time of uniform speed motion from v0 to 0 m / s from s1 to s0; the longitudinal comfort cost is the acceleration of v0 to 0 m / s in the longitudinal displacement time, and if it exceeds a certain range, for example, (-1 m / s2, +1 m / s2), there will be a value of 1.0; the lateral comfort cost is the same, which will not be described herein. The distance from obstacle cost is the distance evaluation of the nearest obstacle obtained by searching (59, s1, l0) in a straight line, and if the distance from the obstacle is very close, a high value can be given to represent collision with the obstacle, and if the distance from the obstacle is very far, a 0 can be given to represent a safe distance from the obstacle.

[0091] It should be noted that the mapping relationship between the above physical meanings and the value of the cost is not a limitation of the content of the present application, but is provided only as an example.

[0092] In this way, the value of each (v, s, l) point at 59 s can be obtained, and the value of each point at 58 s and 57 s can be obtained in the same way, and a four-dimensional graph can be constructed.

[0093] Of course, if there are different parking points, the initial values of different parking points can be determined according to the distances from the ideal target end point. For example, (s0, l0) is the closest, and the initial value is 0, (s0, l1) is more lateral, and the initial value is 10. It should be noted that, similarly, the taking method of this mapping relationship is not a limitation of the content of the present application, but is provided only as an example.

[0094] As a specific embodiment, the step S4 can further refer to 5, as shown in Figure 5 The step S4 specifically includes:

[0095] Step 401: Obtain the values of the points closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph.

[0096] Step 402: Take the values of the points closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as the trajectory evaluation values.

[0097] Step 403: Traverse the evaluation values to obtain the minimum evaluation value.

[0098] Step 404: Confirm the theoretical parking trajectory corresponding to the minimum trajectory value as the actual parking trajectory.

[0099] It should be noted that, for the obtaining of the plurality of theoretical parking trajectories, since the present application is only parking trajectory selection, as a trajectory evaluation method, it does not involve the trajectory generation method. The trajectory generation belongs to the prior art, and this embodiment will not be described here. The plurality of theoretical parking trajectories can be taken as the obtained content as a premise assumption.

[0100] In this embodiment, since the vehicle state is generally given during trajectory planning, including time, vehicle speed, plane coordinates, etc. Thus, only the closest t, the closest v, the closest s, and the closest l of the end of the theoretical parking trajectory need to be found on the aforementioned obtained T-V-S-L four-dimensional graph, i.e. (t, v, s, l), so that the trajectory evaluation value can be found on the T-V-S-L four-dimensional graph. Among them, the theoretical parking trajectory with the lowest evaluation value is the selected actual parking trajectory. The initial value and the transition cost here refer to a positive number greater than 0, and the sum of the initial value and the transition cost is the evaluation value, and the trajectory with the minimum evaluation value is the selected trajectory. Specifically, the values of the points closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph are obtained; the values of the points closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph are taken as the trajectory evaluation values; the minimum evaluation value is obtained by traversing the evaluation values; and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory.

[0101] The embodiment of the present application draws an evaluation area of a parking range according to a terminal position, obtains different parking point positions and corresponding initial reward values in the evaluation area, constructs a retriever in time and position dimensions according to a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment, retrieves a distance between the vehicle and the obstacle, takes a lateral speed change rate and a longitudinal speed change rate of the vehicle as constraints of dynamic programming, takes a lateral acceleration change rate and a longitudinal acceleration change rate of the vehicle and the distance from the obstacle as transition costs, performs multi-terminal dynamic programming on the different parking point positions and the corresponding initial reward values as terminal points of dynamic programming, and constructs a T-V-S-L four-dimensional graph. A value of each point in the T-V-S-L four-dimensional graph is a sum of the initial value and the transition cost. A value of a point closest to an end of a plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph is taken as a trajectory evaluation value, the evaluation values are traversed, and a theoretical parking trajectory corresponding to a minimum evaluation value is confirmed as an actual parking trajectory. In this way, the present application realizes a comprehensive evaluation method for different trajectories facing different terminals by using a multi-terminal dynamic programming algorithm, and comprehensively evaluates different parking trajectories by taking the distance between the parking point and the terminal, the safety degree, and the comfort degree as consideration factors, which is helpful for automatically evaluating the parking trajectories.

[0102] In addition, the embodiment of the present application also provides a parking trajectory selection device, which refers to Figure 6 , and the parking trajectory selection device comprises:

[0103] A division unit 10 is configured to draw an evaluation area of a parking range according to a terminal position, and obtain different parking point positions and corresponding initial values in the evaluation area.

[0104] A first obtaining unit 20 is configured to construct a retriever in time and position dimensions according to a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment, and retrieve a distance between the vehicle and the obstacle.

[0105] A second obtaining unit 30 is configured to take a lateral speed change rate and a longitudinal speed change rate of the vehicle as constraints of dynamic programming, take a lateral acceleration change rate and a longitudinal acceleration change rate of the vehicle and the distance from the obstacle as transition costs, perform multi-terminal dynamic programming on the different parking point positions and the corresponding initial values as terminal points of dynamic programming, construct a T-V-S-L four-dimensional graph, and obtain a value of each point in the T-V-S-L four-dimensional graph.

[0106] A trajectory confirming unit 40 is configured to take a value of a point closest to an end of a plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as a trajectory evaluation value, traverse the evaluation values, and confirm a theoretical parking trajectory corresponding to a minimum evaluation value as an actual parking trajectory.

[0107] In some embodiments of the present application, the process of demarcating the evaluation area of the parking range according to the end position by the demarcating unit 10 can include:

[0108] obtaining the end position;

[0109] obtaining a parking point position at the end position; wherein the parking point position includes a determined parking point position and / or a possible parking point position;

[0110] demarcating a rectangular area as the evaluation area of the parking range according to the parking point position.

[0111] In some embodiments of the present application, the process of obtaining different parking point positions and their corresponding initial values in the evaluation area by the demarcating unit 10 can include:

[0112] obtaining the Manhattan distance between the end position and different parking point positions in the evaluation area;

[0113] taking the square value of each Manhattan distance as the initial value corresponding to the different parking point positions.

[0114] In some embodiments of the present application, the process of constructing the retriever in the time and position dimensions according to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment by the first obtaining unit 20 can include:

[0115] obtaining the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment;

[0116] interpolating the obstacle predicted trajectory to obtain a sampling result under discrete time sampling; wherein the sampling result includes the position and speed of each obstacle;

[0117] constructing the retriever in the time and position dimensions according to the sampling result.

[0118] In some embodiments of the present application, the process of retrieving the distance between the vehicle and the obstacle by the first obtaining unit 20 can include:

[0119] obtaining a first position of the vehicle at a first time;

[0120] obtaining the distance between the vehicle and the obstacle at the first position according to the retriever.

[0121] In some embodiments of the present application, the second acquisition unit 30 performs multi-endpoint dynamic programming by taking the lateral and longitudinal speed change rates of the vehicle as constraints of the dynamic programming, taking the lateral and longitudinal acceleration change rates of the vehicle and the distance to the obstacle as transition costs, taking the different parking point positions and their corresponding initial values as endpoints of the dynamic programming, constructing a T-V-S-L four-dimensional graph, and obtaining the values of each point in the T-V-S-L four-dimensional graph, which can include:

[0122] Taking discrete time as the step of the dynamic programming, and performing multi-endpoint dynamic programming according to time from large to small;

[0123] Taking the lateral and longitudinal speed change rates of the vehicle as constraints of the dynamic programming; taking the lateral and longitudinal acceleration change rates of the vehicle and the distance to the obstacle as transition costs

[0124] Taking the different parking point positions and their corresponding initial values as endpoints of the dynamic programming, constructing a T-V-S-L four-dimensional graph, and obtaining the values of each point in the T-V-S-L four-dimensional graph.

[0125] In some embodiments of the present application, the trajectory confirmation unit 40 takes the values of the points closest to the ends of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as trajectory evaluation values, traverses the evaluation values, and confirms the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory, which can include:

[0126] Obtaining the values of the points closest to the ends of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph;

[0127] Taking the values of the points closest to the ends of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as trajectory evaluation values;

[0128] Traversing the evaluation values to obtain the minimum evaluation value;

[0129] Confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0130] The embodiment of the present application, according to the terminal position demarcated parking range evaluation area, obtains different parking point positions and corresponding initial values in the evaluation area; according to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment, a retriever in time and position dimensions is constructed, and the vehicle-obstacle distance is retrieved; the lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as the constraints of dynamic programming, the lateral acceleration change rate and the longitudinal acceleration change rate of the vehicle and the vehicle-obstacle distance are taken as the transition cost, the different parking point positions and the corresponding first reward values are taken as the dynamic programming endpoints, multi-endpoint dynamic programming is performed, a T-V-S-L four-dimensional graph is constructed, and the values of each point in the T-V-S-L four-dimensional graph are obtained; the values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph are taken as the trajectory evaluation values, the evaluation values are traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory. In this way, the present application realizes a comprehensive evaluation method for different trajectories facing different endpoints by using the multi-endpoint dynamic programming algorithm, and takes the distance between the parking point and the endpoint, the safety degree and the comfort degree as the consideration factors to comprehensively evaluate different parking trajectories, which helps to automatically realize the evaluation of the parking trajectory.

[0131] It should be noted that each unit in the above device can be used to realize each step in the above method, and the corresponding technical effects are achieved, and the present embodiment will not be described here.

[0132] Reference Figure 7 , Figure 7 The structural schematic diagram of the vehicle related to the hardware running environment of the embodiment of the present application.

[0133] As Figure 7 shown, the vehicle can include a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to realize the connection and communication between the components. The user interface 1003 can include a display screen (Display) and an input unit such as a keyboard (Keyboard). The optional user interface 1003 can also include a standard wired interface, a wireless interface. The network interface 1004 can optionally include a standard wired interface, a wireless interface (such as a WI-FI, 4G, 5G interface). The memory 1005 can be a high-speed RAM memory, or a stable memory (non-volatile memory) such as a magnetic disk memory. The memory 1005 can also be an independent storage device from the aforementioned processor 1001.

[0134] Those skilled in the art can understand, Figure 7The structures shown in the figures do not constitute a limitation on the vehicle, which can include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0135] As shown in FIG. 1, the vehicle can include a processor 1001, a memory 1005, a network interface 1004, and a user interface 1003. Figure 7 As shown in FIG. 1, the vehicle can include a processor 1001, a memory 1005, a network interface 1004, and a user interface 1003.

[0136] In the vehicle shown in FIG. 1, the network interface 1004 is mainly used for data communication with an external network; the user interface 1003 is mainly used for receiving input instructions of a user; the vehicle invokes the parking trajectory selection program stored in the memory 1005 through the processor 1001, and performs the following operations: ​ S1: According to the end position, an evaluation area of a parking range is determined, and different parking point positions and their corresponding initial values in the evaluation area are obtained;

[0137] S2: According to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment, a retriever in time and position dimensions is constructed, and the distance between the vehicle and the obstacle is retrieved;

[0138] S3: The lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as constraints of dynamic programming, the lateral acceleration change rate, the longitudinal acceleration change rate, and the distance to the obstacle are taken as transition costs, and the different parking point positions and their corresponding initial values are taken as the end points of dynamic programming for multi-endpoint dynamic programming, a T-V-S-L four-dimensional graph is constructed, and the values of each point in the T-V-S-L four-dimensional graph are obtained;

[0139] S4: The values of the points closest to the end of the theoretical parking trajectory in the T-V-S-L four-dimensional graph are taken as trajectory evaluation values, the evaluation values are traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory.

[0140] Optionally, the step S1 includes:

[0141] Obtaining an end position;

[0142] Obtaining a parking point position at the end position; wherein the parking point position includes a determined parking point position and / or a possible parking point position;

[0143] According to the parking point position, a rectangular area is determined as an evaluation area of a parking range.

[0144] Optionally, the step S1 further includes:

[0145]

[0146] ​Obtaining Manhattan distances between the terminal position and different parking point positions in the evaluation area;

[0147] Taking a square value of each Manhattan distance as an initial value corresponding to a different parking point position.

[0148] Optionally, the step S2 comprises:

[0149] Obtaining a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment;

[0150] Interpolating the obstacle predicted trajectory to obtain a sampling result under discrete time sampling; wherein the sampling result comprises a position and a speed of each obstacle;

[0151] Constructing a retriever in time and position dimensions according to the sampling result.

[0152] Optionally, the step S2 further comprises:

[0153] Obtaining a first position of the vehicle at a first time;

[0154] Obtaining a distance between the vehicle and the obstacle at the first position according to the retriever.

[0155] Optionally, the step S3 comprises:

[0156] C1: taking discrete time as a step length of dynamic programming, and performing multi-terminal dynamic programming according to time from large to small;

[0157] C2: taking a lateral speed change rate and a longitudinal speed change rate of the vehicle as a constraint of dynamic programming; and taking a lateral acceleration change rate, a longitudinal acceleration change rate, and a distance from the vehicle to the obstacle as a transition cost;

[0158] C3: taking the different parking point positions and the initial values corresponding thereto as dynamic programming terminal points, constructing a T-V-S-L four-dimensional graph, and obtaining a value of each point in the T-V-S-L four-dimensional graph.

[0159] Optionally, the step S4 comprises:

[0160] Obtaining a value of a point closest to an end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph;

[0161] Taking the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as a trajectory evaluation value;

[0162] Traversing the evaluation values to obtain a minimum evaluation value;

[0163] Confirming a theoretical parking trajectory corresponding to the minimum evaluation value as an actual parking trajectory.

[0164] According to the embodiment of the present application, the evaluation area of the parking range is determined according to the end position, the different parking point positions and the corresponding initial values in the evaluation area are obtained, the retriever in the time and position dimensions is constructed according to the predicted trajectories of the static obstacles and / or the dynamic obstacles in the vehicle driving environment, the distance between the vehicle and the obstacles is retrieved, the lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as the constraints of the dynamic programming, the lateral acceleration change rate and the longitudinal acceleration change rate of the vehicle and the distance to the obstacles are taken as the transition cost, the different parking point positions and the corresponding initial values are taken as the end points of the dynamic programming, the multi-endpoint dynamic programming is performed, the T-V-S-L four-dimensional graph is constructed, and the values of each point in the T-V-S-L four-dimensional graph are obtained, the values of the points closest to the ends of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph are taken as the trajectory evaluation values, the evaluation values are traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory. In this way, the comprehensive evaluation method of different trajectories facing different end points is realized by the algorithm of the multi-endpoint dynamic programming, the distance between the parking point and the end point, the safety degree, and the comfort degree are taken as the consideration factors to comprehensively evaluate different parking trajectories, and the evaluation of the parking trajectory is automatically realized.

[0165] In addition, the embodiment of the present application also provides a computer readable storage medium, and the computer readable storage medium stores a parking trajectory selection program. When the parking trajectory selection program is executed by a processor, the following operations are realized.

[0166] S1: determining an evaluation area of a parking range according to an end position, and obtaining different parking point positions and corresponding initial values in the evaluation area;

[0167] S2: constructing a retriever in the time and position dimensions according to predicted trajectories of static obstacles and / or dynamic obstacles in a vehicle driving environment, and retrieving a distance between the vehicle and the obstacles;

[0168] S3: taking a lateral speed change rate and a longitudinal speed change rate of the vehicle as constraints of dynamic programming, taking a lateral acceleration change rate and a longitudinal acceleration change rate of the vehicle and the distance to the obstacles as transition cost, performing multi-endpoint dynamic programming by taking the different parking point positions and the corresponding initial values as end points of the dynamic programming, constructing a T-V-S-L four-dimensional graph, and obtaining values of each point in the T-V-S-L four-dimensional graph;

[0169] S4: taking values of points closest to ends of multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as trajectory evaluation values, traversing the evaluation values, and confirming a theoretical parking trajectory corresponding to a minimum evaluation value as an actual parking trajectory.

[0170] Optionally, the step S1 comprises:

[0171] obtaining a terminal position;

[0172] obtaining a parking point position at the terminal position; wherein the parking point position comprises a determined parking point position and / or a possible parking point position;

[0173] delineating a rectangular region as an evaluation region of a parking range according to the parking point position.

[0174] Optionally, the step S1 further comprises:

[0175] obtaining Manhattan distances between the terminal position and different parking point positions in the evaluation region;

[0176] taking a square value of each Manhattan distance as an initial value corresponding to a different parking point position.

[0177] Optionally, the step S2 comprises:

[0178] obtaining a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment;

[0179] interpolating the predicted trajectory of the obstacle to obtain a sampling result under discrete time sampling; wherein the sampling result comprises a position and a speed of each obstacle;

[0180] constructing a retriever in time and position dimensions according to the sampling result.

[0181] Optionally, the step S2 further comprises:

[0182] obtaining a first position of the vehicle at a first time;

[0183] obtaining a distance between the vehicle and the obstacle at the first position according to the retriever.

[0184] Optionally, the step S3 comprises:

[0185] C1: taking discrete time as a step length of dynamic programming, and performing multi-terminal dynamic programming according to time from large to small;

[0186] C2: taking a lateral speed change rate and a longitudinal speed change rate of the vehicle as a constraint of dynamic programming; and taking a lateral acceleration change rate, a longitudinal acceleration change rate, and a distance from the vehicle to the obstacle as a transition cost;

[0187] C3: taking the different parking point positions and corresponding first reward values thereof as dynamic programming terminal points, constructing a T-V-S-L four-dimensional graph, and obtaining a value of each point in the T-V-S-L four-dimensional graph.

[0188] Optionally, the step S4 comprises:

[0189] Obtaining the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph;

[0190] Taking the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as the trajectory evaluation value;

[0191] Traversing the evaluation value to obtain the minimum evaluation value;

[0192] Confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

[0193] In the embodiment of the application, an evaluation area of a parking range is demarcated according to the end position, different parking point positions and corresponding initial values in the evaluation area are obtained, a retriever in the time and position dimensions is constructed according to the predicted trajectories of static obstacles and / or dynamic obstacles in the vehicle driving environment, the distance between the vehicle and the obstacles is retrieved, the lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as the constraints of dynamic programming, the lateral acceleration change rate, the longitudinal acceleration change rate and the distance to the obstacles are taken as the transition cost, the different parking point positions and the corresponding initial values are taken as the end points of dynamic programming, multi-endpoint dynamic programming is performed, a T-V-S-L four-dimensional graph is constructed, and the value of each point in the T-V-S-L four-dimensional graph is obtained, the value of the point closest to the end of the plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph is taken as the trajectory evaluation value, the evaluation value is traversed, the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory. In this way, the comprehensive evaluation method of different trajectories facing different end points is realized by the algorithm of multi-endpoint dynamic programming, the distance between the parking point and the end point, the safety degree and the comfort degree are taken as the consideration factors to comprehensively evaluate different parking trajectories, and this is helpful for automatically realizing the evaluation of the parking trajectory.

[0194] It should be noted that, in this document, the terms "comprise", "comprising", or any other variant thereof are intended to cover non-exclusive inclusions, such that processes, methods, articles, or systems that comprise a list of elements do not only include those elements, but can also include other elements that are not expressly listed, or other elements inherent to such processes, methods, articles, or systems. Without more limitations, the element defined by the sentence "comprises a" does not exclude the presence of additional identical elements in the process, method, article, or system that includes the element.

[0195] The above-mentioned embodiment numbers of the application are only for description, and do not represent the advantages and disadvantages of the embodiments.

[0196] Those skilled in the art can clearly understand the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, also can be through hardware, but in many cases the former is the better embodiment. Based on such understanding, the technical solutions of the present application essentially or say the part of the prior art contribution can be embodied in the form of software products, the computer software product is stored in a storage medium (such as ROM / RAM, magnetic disc, optical disc) as described above, including a number of instructions to make a terminal device (may be a mobile phone, computer, server, controller, or network equipment, etc.) executes the method described in various embodiments of the present application.

[0197] The above is only the preferred embodiment of the present application, not therefore limit the patent scope of the present application, any equivalent structure or equivalent flow transformation made by using the content of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A parking trajectory selection method characterized by, The method comprises the following steps: S1: According to the end position, the evaluation area of the parking range is drawn, and different parking point positions and their corresponding initial values in the evaluation area are obtained; S2: According to the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment, a retriever in time and position dimensions is constructed, and the distance between the vehicle and the obstacle is retrieved; S3: The lateral speed change rate and the longitudinal speed change rate of the vehicle are taken as the constraints of dynamic programming, the lateral acceleration change rate and the longitudinal acceleration change rate of the vehicle and the distance from the vehicle to the obstacle are taken as the transition cost, and the different parking point positions and their corresponding initial values are taken as the end points of dynamic programming for multi-endpoint dynamic programming, so as to construct a T-V-S-L four-dimensional graph and obtain the values of each point in the T-V-S-L four-dimensional graph; S4: The values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph are taken as the trajectory evaluation values, the evaluation values are traversed, and the theoretical parking trajectory corresponding to the minimum evaluation value is confirmed as the actual parking trajectory.

2. The parking trajectory selection method according to claim 1, characterized in that: The step S1 comprises: obtaining an end position; obtaining a parking point position at the end position; wherein the parking point position comprises a determined parking point position and / or a possible parking point position; drawing a rectangular area as an evaluation area of a parking range according to the parking point position.

3. The parking trajectory selection method according to claim 2, characterized in that, The step S1 further comprises: obtaining the Manhattan distance between the end position and different parking point positions in the evaluation area; taking the square value of each Manhattan distance as the initial value corresponding to the different parking point positions.

4. The parking trajectory selection method according to claim 1, characterized by, The step S2 comprises: obtaining the predicted trajectory of the static obstacle and / or dynamic obstacle in the vehicle driving environment; interpolating the obstacle predicted trajectory to obtain a sampling result under discrete time sampling; wherein the sampling result comprises the position and speed of each obstacle; constructing a retriever in time and position dimensions according to the sampling result.

5. The parking trajectory selection method according to claim 4, characterized in that, The step S2 further comprises: obtaining a first position of the vehicle at a first time; obtaining the distance between the vehicle and the obstacle at the first position according to the retriever.

6. The parking trajectory selection method according to claim 1, characterized by, The step S3 comprises: C1: taking discrete time as the step length of dynamic programming, and performing multi-endpoint dynamic programming according to time from large to small; C2: taking the lateral speed change rate and the longitudinal speed change rate of the vehicle as the constraints of dynamic programming; taking the lateral acceleration change rate and the longitudinal acceleration change rate of the vehicle and the distance from the vehicle to the obstacle as the transition cost; C3: taking the different parking point positions and their corresponding first reward values as the end points of dynamic programming, constructing a T-V-S-L four-dimensional graph and obtaining the values of each point in the T-V-S-L four-dimensional graph.

7. The parking trajectory selection method according to claim 1, characterized by, The step S4 comprises: obtaining the values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph; taking the values of the points closest to the end of the multiple theoretical parking trajectories in the T-V-S-L four-dimensional graph as the trajectory evaluation values; traversing the evaluation values to obtain the minimum evaluation value; confirming the theoretical parking trajectory corresponding to the minimum evaluation value as the actual parking trajectory.

8. A parking trajectory selection apparatus characterized by comprising: The parking trajectory selection device comprises: The division unit is configured to divide an evaluation area of a parking range according to a terminal position, and obtain different parking point positions and corresponding initial values in the evaluation area; The first obtaining unit is configured to construct a retriever in time and position dimensions according to a predicted trajectory of a static obstacle and / or a dynamic obstacle in a vehicle driving environment, and retrieve a distance between the vehicle and the obstacle; The second obtaining unit is configured to perform multi-endpoint dynamic programming by taking a lateral speed change rate and a longitudinal speed change rate of the vehicle as a constraint of dynamic programming, taking lateral acceleration change rate, longitudinal acceleration change rate and the distance between the vehicle and the obstacle as a transition cost, and taking the different parking point positions and the corresponding initial values as dynamic programming endpoints, to construct a T-V-S-L four-dimensional graph and obtain a value of each point in the T-V-S-L four-dimensional graph; The trajectory confirmation unit is configured to take a value of a point closest to an end of a plurality of theoretical parking trajectories in the T-V-S-L four-dimensional graph as a trajectory evaluation value, traverse the evaluation value, and confirm a theoretical parking trajectory corresponding to a minimum evaluation value as an actual parking trajectory.

9. A vehicle, characterized by The vehicle comprises a memory, a processor, and a parking trajectory selection program stored in the memory and executable on the processor, and the parking trajectory selection program is configured to implement steps of the parking trajectory selection method according to any one of claims 1 to 7.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements steps of the parking trajectory selection method according to any one of claims 1 to 7.

Images