Running corridor generation method and equipment based on reachable set, and storage medium

Abstract

The invention discloses a driving corridor generation method and device based on a reachable set and a storage medium, and the method comprises the steps: carrying out the initialization of the reachable set according to the initial state information of a vehicle, the position and shape information of an obstacle, the geometric information of a road, the planning propagation time, and the state propagation interval time; according to the change of the real-time motion state information in the vehicle motion process, the reachable set state at each moment is updated in a propagation mode; establishing a graph model according to the reachable set at each moment; performing deep search in the graph model to obtain a plurality of traveling galleries; and on the basis of the reachable set at each moment, through a plurality of evaluation indexes, selecting an optimal driving corridor in the plurality of driving corridors. On the basis of the reachable set, time and space are combined, the drivable corridor is directly generated, and the algorithm is short in consumed time and high in efficiency. And the solution space completeness of time and space combined mapping ensures the completeness of the algorithm, so that the corridor generation failure is avoided, and the intelligent driving level is integrally improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of convex space decision-making for intelligent vehicles, and specifically relates to a method, device and storage medium for generating a driving corridor based on a reachable set. Background Art

[0002] The main application scenarios of autonomous driving technology include: smart city travel, intelligent sanitation, unmanned logistics, Sinotruk mining trucks, etc., and autonomous driving (Autonomous Driving) mainly includes: perception, positioning, prediction, decision-making, planning, control and other modules.

[0003] At present, the generation of autonomous driving corridors is mainly based on search and expansion solutions. First, a reference line is searched and then expanded to both sides to obtain the driving corridor. Among them, search refers to searching for a reference line in the graph model based on specific preference rules, such as Figure 1 Extension means extending the travel corridor to both sides based on an initial reference line. In the Flechet coordinate system, Figure 2 shown.

[0004] Through dynamic programming (DP) graph search, the basic process of generating driving corridors is generated, such as Figure 3 As shown, it specifically includes the following steps:

[0005] (1) Use static information (high-precision map) to establish the current DP map. The vehicle, obstacles, and some prior information are projected into the DP map. The DP grids occupied by obstacles are non-drivable areas.

[0006] (2) Refresh the dp map weight based on information such as road centerline, road boundary, obstacle grid, target point grid, and historical reference line (if any).

[0007] (3) Starting from the projection point of the rear axle center of the vehicle, traverse the dp graph and search for a minimum loss reference line to the target point.

[0008] (4) Each dp grid on the reference line extends to both sides and stops when it hits a boundary or obstacle, thus obtaining a drivable corridor.

[0009] The above graph search method can ensure global optimality, but the granularity of establishing the graph model is a key issue. If the step size is too large, there will be a problem of no solution space. If the step size is too small, it will lead to low algorithm efficiency and high time consumption. Summary of the invention

[0010] In order to solve the above technical problems, the present invention proposes a method, device and storage medium for generating a driving corridor based on a reachable set.

[0011] In order to achieve the above object, the technical solution of the present invention is as follows:

[0012] In a first aspect, the present invention discloses a method for generating a travel corridor based on a reachable set, comprising:

[0013] Step S1: Initialize the reachable set according to the vehicle's initial state information, obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time;

[0014] Step S2: According to the changes in the real-time motion state information during the vehicle movement, the reachable set state at each moment is updated by propagation;

[0015] Step S3: Establish a graph model based on the reachable set at each moment;

[0016] Step S4: performing a deep search in the graph model to obtain multiple driving corridors;

[0017] Step S5: Based on the reachable set at each moment, the optimal driving corridor is selected from multiple driving corridors through a variety of evaluation indicators.

[0018] Based on the above technical solution, the following improvements can be made:

[0019] As a preferred solution, step S2 specifically includes:

[0020] Step S2.1: Based on the motion state information of the vehicle at each reachable point in the reachable set at the previous moment, calculate all the spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position, and propagate the reachable set;

[0021] Step S2.2: Discretize the reachable set;

[0022] Step S2.3: Combining the reachable points with the same spatial position and different motion state information obtained by propagating multiple reachable points into a reachable point corresponding to the motion state information;

[0023] Step S2.4: Use the scanning line algorithm to vertically divide the reachable sets so that each reachable set is convex in space;

[0024] Step S2.5: perform collision detection on each possible spatial position, and remove the reachable points with collision detection from the reachable set;

[0025] Step S2.7: Determine whether the current time reaches the planned transmission time;

[0026] If so, save the reachable set at each moment;

[0027] Otherwise, continue updating the reachable set.

[0028] As a preferred solution, step S2 further includes:

[0029] Step S2.6: Determine whether there is a collision between the current reachable set and the obstacle;

[0030] If so, the reachable set is divided along the short axis of the horizontal axis and the vertical axis until the part that collides with the obstacle is deleted;

[0031] Otherwise, no action is taken.

[0032] As a preferred solution, in step S5, the evaluation index includes: one or more of space evaluation, safety evaluation and connectivity evaluation.

[0033] In a second aspect, the present invention further discloses a travel corridor generation device based on a reachable set, comprising:

[0034] An initialization module is used to initialize the reachable set according to the vehicle's initial state information, obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time;

[0035] The reachable set update module is used to update the reachable set state at each moment by propagating the changes in the real-time motion state information during the vehicle movement;

[0036] A graph model building module, used to build a graph model based on the reachable set at each moment;

[0037] A deep search module is used to perform deep search in the graph model to obtain multiple driving corridors;

[0038] The selection module is used to select the optimal driving corridor from multiple driving corridors based on the reachable set at each moment through multiple evaluation indicators.

[0039] As a preferred solution, the reachable set update module includes:

[0040] A propagation unit is used to calculate all spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position according to the motion state information of the vehicle at each reachable point in the reachable set at the previous moment, and propagate the reachable set;

[0041] Discrete units, used to discretize reachable sets;

[0042] A combining unit, used for combining reachable points with the same spatial position and different motion state information obtained by propagating multiple reachable points into a reachable point with corresponding motion state information;

[0043] A vertical segmentation unit is used to vertically segment the reachable sets using a scanning line algorithm, so that each reachable set is convex in space;

[0044] A collision detection unit is used to perform collision detection on each possible spatial position and delete the reachable points with collision detection from the reachable set;

[0045] A judgment unit, used to judge whether the current time reaches the planned transmission time;

[0046] If so, save the reachable set at each moment;

[0047] Otherwise, continue updating the reachable set.

[0048] As a preferred solution, the reachable set update module also includes:

[0049] The segmentation unit is used to segment the reachable set along the short axis of the horizontal axis and the vertical axis when the current reachable set collides with the obstacle, until the part that collides with the obstacle is deleted.

[0050] As a preferred solution, the evaluation indicators include: one or more of space evaluation, safety evaluation and connectivity evaluation.

[0051] In addition, in a third aspect, the present invention further discloses a storage medium, which stores one or more computer-readable programs, and the one or more programs include instructions, and the instructions are suitable for being loaded by a memory and executing the above-mentioned driving corridor generation method.

[0052] The present invention discloses a method, device and storage medium for generating a travel corridor based on a reachable set, which has the following beneficial effects:

[0053] First, the present invention combines time and space based on reachable sets to directly generate drivable corridors, and the algorithm is time-saving and efficient.

[0054] Second, the integrity of the solution space of the joint temporal and spatial mapping ensures the completeness of the algorithm, thereby avoiding the failure of corridor generation and improving the overall level of intelligent driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0056] Figure 1 A reference line schematic diagram is obtained by searching in the graph model provided by the prior art.

[0057] Figure 2 A schematic diagram of a corridor is obtained by extending both sides of a reference line in a Flechet coordinate system provided by the prior art.

[0058] Figure 3 The basic flow chart of the driving corridor is generated by dynamic planning provided by the existing technology.

[0059] Figure 4 A flowchart of a method for generating a driving corridor provided in an embodiment of the present invention.

[0060] Figure 5 A reachable set update flow chart provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0061] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] In addition, the expression of “comprising” an element is an “open” expression, which merely means that corresponding components or steps exist, and should not be interpreted as excluding additional components or steps.

[0064] The method of generating drivable space based on dynamic programming search can only achieve the goal of neither losing solution space nor being inefficient and time-consuming under ideal scenarios, accurate mapping information, and appropriate mapping step length. However, in actual situations, dynamic programming step length sampling is difficult to determine, and short step lengths are generally used to ensure solution space, which is a compromise of low efficiency and high time consumption.

[0065] How to reduce algorithm time consumption and improve algorithm efficiency while ensuring the solution space is a difficult problem faced in generating drivable space. Through analysis, the dynamic programming algorithm mainly consumes time in horizontal and vertical sampling and mapping, and updating loss items. The present invention proposes a method, device and storage medium for generating a driving corridor based on a reachable set.

[0066] In order to achieve the purpose of the present invention, in an embodiment of a travel corridor generation method based on a reachable set, as shown in FIG. Figure 4 As shown in FIG. 1 , the driving corridor generation method includes:

[0067] Step S1: Initialize the reachable set according to the vehicle's initial state information (such as starting position, direction, speed, acceleration interval, etc.), obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time;

[0068] Step S2: According to the changes in the real-time motion state information (such as the position, direction, speed, etc. of the vehicle) during the vehicle movement, the reachable set state at each moment is updated by propagation;

[0069] Step S3: Establish a graph model based on the reachable set at each moment;

[0070] Step S4: performing a deep search in the graph model to obtain multiple driving corridors;

[0071] Step S5: Based on the reachable set at each moment, the optimal driving corridor is selected from multiple driving corridors through a variety of evaluation indicators.

[0072] Further, each step is described in detail below.

[0073] Step S1 is mainly to initialize the reachable set. The initialization operation is only performed once when the initialization is planned.

[0074] like Figure 5 As shown, step S2 is a step of updating the reachable set, which specifically includes steps S2.1 to S2.7.

[0075] Step S2.1: Based on the motion state information of the vehicle at each reachable point in the reachable set at the previous moment (such as vehicle position, direction, speed, acceleration range, etc.), calculate all the spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position (such as vehicle posture, speed, acceleration state, etc.), and propagate the reachable set.

[0076] It is worth noting that points with the same spatial coordinates in the reachable set may be propagated from multiple reachable points in the previous state. The reachable points of all states are recorded to ensure the integrity of the solution space.

[0077] During the propagation process, if the vehicle reaches the set maximum speed, the upper speed bound will no longer be updated. To ensure that the reachable set propagation does not oscillate and the forward search state is stable, the lower speed bound can be set to 0.

[0078] Step S2.2: Discretize the reachable set.

[0079] The current reachable set contains all the reachable points in the reachable set of the previous moment without considering obstacles. When propagating and expanding, it is calculated from the x and y directions and then obtained through the Cartesian product.

[0080] In order to avoid huge data volume and few effective points, discretization is used to process the reachable set to ensure the stability of the reachable set in the propagation process.

[0081] Step S2.3: Under the premise of ensuring the integrity of the solution space, the reachable points with the same spatial position and different speed states obtained by propagating multiple reachable points are combined into a reachable point with one speed state.

[0082] Step S2.4: Use the scanning line algorithm to vertically divide the reachable sets so that each reachable set is convex in space.

[0083] Step S2.5: Perform collision detection on each possible spatial position to eliminate the risk of collision with obstacles, and delete the reachable points with collision detection from the reachable set.

[0084] The collision points should not be propagated in the next solution space, and the pruning idea can further reduce the algorithm time consumption.

[0085] Step S2.6: Determine whether there is a collision between the current reachable set and the obstacle;

[0086] If so, the reachable set is split along the short axis of the horizontal axis and the vertical axis until the reachable set is small enough to delete the reachable subset that still collides with the obstacle;

[0087] Otherwise, no action is taken.

[0088] Vertically splitting the reachable set can ensure that the algorithm remains complete after pruning.

[0089] Step S2.7: Determine whether the current time reaches the planned transmission time;

[0090] If so, save the reachable set at each moment;

[0091] Otherwise, continue updating the reachable set.

[0092] Step S3 builds a graph model based on the reachable set. The reachable set at each moment is obtained by propagation. The reachable sets are connected in space and time, and the graph model is naturally constructed.

[0093] Step S4 performs a deep search (DFS) in the graph model established by the reachable set to obtain multiple driving corridors.

[0094] In step S5, based on the reachable set at each moment, the optimal driving corridor is selected from multiple driving corridors through a variety of evaluation indicators.

[0095] Evaluation indicators include: spatial evaluation, safety evaluation and connectivity evaluation. Spatial evaluation: whether the reachable set propagation contains the topology of key nodes, the distance between the final state of the reachable set and the destination of the driving target. Safety evaluation: the risk of collision with obstacles during the reachable set propagation process, the number of reachable set divisions accumulated due to collision judgment. Connectivity evaluation: whether there is a decay in the propagation state of the reachable set, the approximation of the fitting quadratic programming, etc.

[0096] The present invention constructs the map in a reachable set propagation manner, and the non-horizontal and vertical sampling manner effectively reduces the algorithm time consumption. In addition, the non-horizontal and vertical sampling manner more effectively explores the solution space and reduces limitations, and the present invention is complete.

[0097] In summary, the present invention infers the range of changes in the vehicle's motion state from the perspective of the union of time and space, achieving real-time planning while ensuring the integrity of the solution space. The present invention deduces the reachable set space according to the increment of time. The reachable set space before and after the deduction is connected, and the establishment of the graph model is efficient and time-saving. The present invention uses the reachable set as a state representation of the vehicle's driving process, which can efficiently reflect the vehicle's state changes and carry more driving information for establishing the graph model and selecting the final optimal driving corridor.

[0098] In addition, in an embodiment of the driving corridor generation device based on the reachable set, the driving corridor generation device includes:

[0099] An initialization module is used to initialize the reachable set according to the vehicle's initial state information, obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time;

[0100] The reachable set update module is used to update the reachable set state at each moment by propagating the changes in the real-time motion state information during the vehicle movement;

[0101] A graph model building module, used to build a graph model based on the reachable set at each moment;

[0102] A deep search module is used to perform deep search in the graph model to obtain multiple driving corridors;

[0103] The selection module is used to select the optimal driving corridor from multiple driving corridors based on the reachable set at each moment through multiple evaluation indicators.

[0104] Furthermore, the reachable set update module includes:

[0105] A propagation unit is used to calculate all spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position according to the motion state information of the vehicle at each reachable point in the reachable set at the previous moment, and propagate the reachable set;

[0106] Discrete units, used to discretize reachable sets;

[0107] A combining unit, used for combining reachable points with the same spatial position and different motion state information obtained by propagating multiple reachable points into a reachable point with corresponding motion state information;

[0108] A vertical segmentation unit is used to vertically segment the reachable sets using a scanning line algorithm, so that each reachable set is convex in space;

[0109] A collision detection unit is used to perform collision detection on each possible spatial position and delete the reachable points with collision detection from the reachable set;

[0110] A judgment unit, used to judge whether the current time reaches the planned transmission time;

[0111] If so, save the reachable set at each moment;

[0112] Otherwise, continue updating the reachable set.

[0113] Furthermore, the reachable set update module also includes:

[0114] The segmentation unit is used to segment the reachable set along the short axis of the horizontal axis and the vertical axis when the current reachable set collides with the obstacle, until the part that collides with the obstacle is deleted.

[0115] Furthermore, the evaluation indicators include: one or more of space evaluation, safety evaluation and connectivity evaluation.

[0116] In this embodiment, the specific content of the driving corridor generation device based on the reachable set is similar to the content of the driving corridor generation method based on the reachable set disclosed in the above embodiment, and will not be repeated here.

[0117] In an embodiment of the storage medium, the storage medium stores one or more computer-readable programs, and the one or more programs include instructions, and the instructions are suitable for being loaded by the memory and executing the above-mentioned driving corridor generation method.

[0118] The present invention discloses a method, device and storage medium for generating a travel corridor based on a reachable set, which has the following beneficial effects:

[0119] First, the present invention combines time and space based on reachable sets to directly generate drivable corridors, and the algorithm is time-saving and efficient.

[0120] Second, the integrity of the solution space of the joint temporal and spatial mapping ensures the completeness of the algorithm, thereby avoiding the failure of corridor generation and improving the overall level of intelligent driving.

[0121] It should be understood that the various techniques described herein can be implemented in combination with hardware or software, or a combination thereof. Thus, the method and apparatus of the present invention, or some aspects or portions of the method and apparatus of the present invention, can take the form of program codes (i.e., instructions) embedded in a tangible medium, such as a floppy disk, a CD-ROM, a hard disk drive, or any other machine-readable storage medium, wherein when the program is loaded into a machine such as a computer and executed by the machine, the machine becomes a device for practicing the present invention.

[0122] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for illustrating the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which shall fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A travel corridor generation method based on reachable sets, characterized in that: include: Step S1: Initialize the reachable set according to the vehicle's initial state information, obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time; Step S2: According to the changes in the real-time motion state information during the vehicle movement, the reachable set state at each moment is updated by propagation; Step S3: Establish a graph model based on the reachable set at each moment; Step S4: performing a deep search in the graph model to obtain multiple driving corridors; Step S5: Based on the reachable set at each moment, the optimal driving corridor is selected from multiple driving corridors through a variety of evaluation indicators.

2. The driving corridor generation method according to claim 1, characterized in that: The step S2 specifically includes: Step S2.1: Based on the motion state information of the vehicle at each reachable point in the reachable set at the previous moment, calculate all the spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position, and propagate the reachable set; Step S2.2: discretize the reachable set; Step S2.3: Combining the reachable points with the same spatial position and different motion state information obtained by propagating multiple reachable points into a reachable point corresponding to the motion state information; Step S2.4: Use the scanning line algorithm to vertically divide the reachable sets so that each reachable set is convex in space; Step S2.5: perform collision detection on each possible spatial position, and remove the reachable points with collision detection from the reachable set; Step S2.7: Determine whether the current time reaches the planned transmission time; If so, save the reachable set at each moment; Otherwise, continue updating the reachable set.

3. The driving corridor generation method according to claim 2, characterized in that: The step S2 further comprises: Step S2.6: Determine whether there is a collision between the current reachable set and the obstacle; If so, the reachable set is divided along the short axis of the horizontal axis and the vertical axis until the part that collides with the obstacle is deleted; Otherwise, no action is taken.

4. The method for generating a driving corridor according to claim 1, characterized in that: In step S5, the evaluation index includes one or more of space evaluation, safety evaluation and connectivity evaluation.

5. A travel corridor generation device based on a reachable set, characterized in that: include: An initialization module is used to initialize the reachable set according to the vehicle's initial state information, obstacle position and shape information, road geometry information, planning propagation time, and state propagation interval time; The reachable set update module is used to update the reachable set state at each moment by propagating the changes in the real-time motion state information during the vehicle movement; A graph model building module, used to build a graph model based on the reachable set at each moment; A deep search module is used to perform deep search in the graph model to obtain multiple driving corridors; The selection module is used to select the optimal driving corridor from multiple driving corridors based on the reachable set at each moment through multiple evaluation indicators.

6. The driving corridor generation device according to claim 5, characterized in that: The reachable set update module includes: A propagation unit is used to calculate all spatial positions that the vehicle can reach within a state propagation interval and the motion state information of the vehicle that reaches the spatial position according to the motion state information of the vehicle at each reachable point in the reachable set at the previous moment, and propagate the reachable set; Discrete units, used to discretize reachable sets; A combining unit, used for combining reachable points with the same spatial position and different motion state information obtained by propagating multiple reachable points into a reachable point with corresponding motion state information; A vertical segmentation unit is used to vertically segment the reachable sets using a scanning line algorithm, so that each reachable set is convex in space; A collision detection unit is used to perform collision detection on each possible spatial position and delete the reachable points with collision detection from the reachable set; A judgment unit, used to judge whether the current time reaches the planned transmission time; If so, save the reachable set at each moment; Otherwise, continue updating the reachable set.

7. The driving corridor generation device according to claim 6, characterized in that: The reachable set update module also includes: The segmentation unit is used to segment the reachable set along the short axis of the horizontal axis and the vertical axis when the current reachable set collides with the obstacle, until the part that collides with the obstacle is deleted.

8. The driving corridor generating device according to claim 6, characterized in that: The evaluation indicators include: one or more of space evaluation, safety evaluation and connectivity evaluation.

9. A storage medium, characterized in that The storage medium stores one or more computer-readable programs, and the one or more programs include instructions, and the instructions are suitable for being loaded by the memory and executing the driving corridor generation method described in any one of claims 1 to 4.

Images