CN112824198A

CN112824198A - Trajectory decision method, apparatus, device and storage medium

Info

Publication number: CN112824198A
Application number: CN201911151252.3A
Authority: CN
Inventors: 李柏
Original assignee: Beijing Jingdong Qianshi Technology Co Ltd
Current assignee: Beijing Jingdong Qianshi Technology Co Ltd
Priority date: 2019-11-21
Filing date: 2019-11-21
Publication date: 2021-05-21
Anticipated expiration: 2039-11-21
Also published as: CN112824198B

Abstract

The embodiment of the invention discloses a track decision method, a device, equipment and a storage medium, wherein the method comprises the following steps: acquiring a road passing area of a structured road to be driven by a target vehicle; determining candidate track information of each candidate track in a road passing area, wherein the candidate track information comprises path information and preset parameter information, and the preset parameter comprises time information or speed information; and determining a target track of the running of the target vehicle from the candidate tracks according to the obstacle distribution data and the candidate track information in the road passing area. By the technical scheme of the embodiment of the invention, the accuracy of the track decision can be improved.

Description

Trajectory decision method, apparatus, device and storage medium

Technical Field

The embodiment of the invention relates to an automatic driving technology, in particular to a track decision method, a device, equipment and a storage medium.

Background

Typically, vehicles may be autonomous in a structured road. Trajectory decisions (i.e., running decisions) are often made prior to planning the trajectory of the vehicle to determine which side of the vehicle is around the obstacle, and whether to preempt or give way.

At present, a path + speed decision mode is often used for trajectory decision, and the specific decision process is as follows: firstly, a driving path is decided, then a corresponding speed is matched based on the driving path, so that a rough driving track is decided, and then the driving track is subjected to smoothing processing and the like during track planning.

However, in the process of implementing the present invention, the inventor finds that at least the following problems exist in the prior art:

in the existing route + speed decision mode, when a route is decided, no factor related to speed is considered, namely the situation of dynamic obstacles is not considered, and subsequently, when the speed is decided, the situation that a vehicle cannot reach a destination due to the fact that the number of dynamic obstacles existing on the decided route is large and the reasonable speed cannot be matched possibly occurs, so that a reasonable and feasible decision track cannot be obtained, and the accuracy of track decision is greatly reduced.

Disclosure of Invention

The embodiment of the invention provides a track decision method, a track decision device, track decision equipment and a storage medium, which are used for improving the accuracy of track decision.

In a first aspect, an embodiment of the present invention provides a trajectory decision method, including:

acquiring a road passing area of a structured road to be driven by a target vehicle;

determining candidate track information of each candidate track in the road passing area, wherein the candidate track information comprises path information and preset parameter information, and the preset parameter comprises time information or speed information;

and determining a target track of the running of the target vehicle from each candidate track according to the obstacle distribution data and each candidate track information in the road passing area.

In a second aspect, an embodiment of the present invention further provides a trajectory decision device, including:

the road passing area acquisition module is used for acquiring a road passing area of a structured road to be driven by a target vehicle;

a candidate track information determining module, configured to determine candidate track information of each candidate track in the road passing area, where the candidate track information includes path information and preset parameter information, and the preset parameter includes time information or speed information;

and the target track determining module is used for determining a target track of the running of the target vehicle from each candidate track according to the obstacle distribution data in the road passing area and the information of each candidate track.

In a third aspect, an embodiment of the present invention further provides an apparatus, where the apparatus includes:

one or more processors;

a memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement a trajectory decision method as provided by any of the embodiments of the present invention.

In a fourth aspect, the embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the trajectory decision method provided in any embodiment of the present invention.

According to the embodiment of the invention, the candidate track information of each candidate track in the road passing area of the structured road is determined, and the candidate track information comprises the path information and the preset parameter information, namely the time information or the speed information, so that when the target track of the running target vehicle is determined according to the obstacle distribution data and the candidate track information in the road passing area, the track decision can be simultaneously carried out on the basis of the factors of the path, the speed or the time at one time, the more accurate, reasonable and feasible target track is obtained, and the decision accuracy is greatly improved.

Drawings

Fig. 1 is a flowchart of a trajectory decision method according to an embodiment of the present invention;

FIG. 2 is an example of determining candidate trajectories according to an embodiment of the present invention;

fig. 3 is a flowchart of a trajectory decision method according to a second embodiment of the present invention;

fig. 4 is a flowchart of a trajectory decision method according to a third embodiment of the present invention;

fig. 5 is a schematic structural diagram of a trajectory decision device according to a fourth embodiment of the present invention;

fig. 6 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Example one

Fig. 1 is a flowchart of a trajectory decision method according to an embodiment of the present invention, which is applicable to a situation of performing trajectory decision on an automatically driven vehicle on a structured road. The method may be performed by a trajectory decision means, which may be implemented in software and/or hardware, integrated in a device with autonomous driving functions, such as any type of vehicle. As shown in fig. 1, the method specifically includes the following steps:

and S110, acquiring a road passing area of the structured road to be driven by the target vehicle.

The target vehicle may refer to any type of vehicle currently to be subjected to trajectory decision making. The structured road may refer to a road having a driving rule on which the target vehicle is to be driven. For example, the structured road may be a road having a guide line to indicate a driving direction, such as a city road or the like. The road passing area may refer to a passable area range in the structured road, and may be a road area surrounded by left and right boundaries of the structured road.

S120, determining the information of the tracks to be selected of all the tracks to be selected in the road passing area, wherein the information of the tracks to be selected comprises path information and preset parameter information, and the preset parameters comprise time information or speed information.

The candidate trajectory may be a trajectory that the target vehicle can travel in a road passing area. The preset parameter may be a parameter for directly or indirectly characterizing the running speed condition of the target vehicle, and may be a speed parameter or a time parameter. The candidate trajectory information comprises path information used for representing a running route of the target vehicle and preset parameter information used for representing a running speed condition of the target vehicle, so that when trajectory decision is made based on the candidate trajectory information, factors of the path, the speed or the time can be considered at one time, and the trajectory decision is more accurate.

Specifically, all the candidate tracks in the road passing area and the candidate track information of each candidate track can be determined based on a preset layered sampling mode. Exemplarily, S120 may include: according to the preset number of layers, layering the guiding lines in the road traffic area, and taking the normal line at the intersection point position of each layer and the guiding line as the boundary line of the corresponding layer; dividing the boundary of each layer according to the number of preset nodes, and taking each divided node as a running node of the target vehicle on the boundary; determining each parameter value to be selected when the target vehicle runs to a running node according to the preset parameter value quantity; and respectively selecting a driving node on the boundary line of each layer, respectively selecting a parameter value to be selected for each selected driving node, and taking each selected driving node and each parameter value to be selected as the track information to be selected of a track to be selected.

The preset number of layers can be preset, and the total number of layers divided by the guiding lines in the road passing area is the preset number. The preset number of nodes may be a total number of travel nodes divided on a boundary line of each floor, which is set in advance. The travel node may refer to a position where the vehicle travels onto the boundary line. The number of values of the preset parameter can be preset, and the total number of all the values of the preset parameter can be preset in the maximum allowable range.

Specifically, in the road passing area, the guiding lines in the road passing area may be layered into a preset number of layers at equal intervals or unequal intervals according to mileage, and a normal line of the guiding line is constructed at an intersection position of each layer and the guiding line, and the normal line is used as a boundary line of the corresponding layer. The division into the preset node number is performed at equal intervals or unequal intervals on the boundary of each layer, and each divided node is used as a running node of the target vehicle on the boundary. When the preset parameter is a speed parameter, that is, the average longitudinal speed of the target vehicle moving from the current floor to the next floor, values whose number is the number of values of the preset parameter may be set for the average longitudinal speed based on the maximum allowable longitudinal speed. When the preset parameter is a time parameter, that is, a time interval consumed for the target vehicle to move from the current floor to the next floor, values whose number is the number of values of the preset parameter may be set for the time interval based on the length (i.e., the mileage) of the guidance line and the maximum allowable longitudinal speed. Where longitudinal refers to the direction along the leader line and the corresponding transverse refers to the direction perpendicular to the leader line. The method comprises the steps of selecting a driving node from all driving nodes on the boundary line of each layer, selecting a parameter value to be selected from all parameter values to be selected for each selected driving node, taking each selected driving node and each parameter value to be selected as the track information to be selected of a track to be selected, and performing the steps in the same manner until the track information to be selected of all different tracks to be selected is determined.

Illustratively, fig. 2 gives an example of determining the candidate trajectory. The guiding lines in fig. 2 can be expressed in a cartesian coordinate system (i.e., a rectangular coordinate system) as: Γ (x(s), y (s)), where s ∈ [0, L ∈]Corresponding to the longitudinal mileage interval involved in the trajectory decision. The section of guiding line can be divided into a preset number of layers N at equal intervals according to mileage_layerThe mileage corresponding to the k-th layer is

The intersection position of the k layer and the finger line is P_k＝(x(s_k),y(s_k) And P) on the guideline_kThe normal at (b) is taken as the boundary k of the kth layer. Setting a preset node number N on the boundary line of each layer_nodeAnd the driving nodes are equally spaced. The starting point position of the target vehicle may be set to Q₀This may correspond to a travel node on layer 0 (x (0), y (0)). By selecting a respective driving node Q on each level_kAnd sequentially connecting the selected running nodes based on the sequence of layers to obtain a broken line segment

As a complete candidate path in a cartesian coordinate system.

When the preset parameter is the average longitudinal speed of the target vehicle moving from the current layer to the next layer, the average longitudinal speed of each node on the k layer

Is set of values of

Comprises the following steps:

wherein v is_maxRepresents the maximum longitudinal speed allowed; n is a radical of_parThe method refers to the number of values of preset parameters. When j is equal to 0, the value of j,

0, indicating that the target vehicle does not reach the k-th floor, which is farthest to reach the k-1 th floor, and as far as the target vehicle can reach the k-1 th floor, the average longitudinal velocity from the front floor is required

And (6) determining.

Similarly, when the preset parameter is the time interval consumed by the target vehicle to move from the current layer to the next layer, the set Ψ formed by the values of the time interval Δ t of each node on the k-th layer_tComprises the following steps:

when j is equal to 0, the value of j,

is 0, i.e., Δ t tends to infinity, indicating that the target vehicle does not reach the k-th layer, which reaches the k-1 st layer farthestAs for the target vehicle to be able to reach the k-1 st floor, it needs to be determined by the time interval Δ t of the front floor.

It should be noted that, because the driving nodes on each layer and the parameter values to be selected of each driving node are independently selected, the driving nodes and the parameter values to be selected of each driving node can be obtained

And candidate track information of different candidate tracks.

S130, determining a target track of the running of the target vehicle from the candidate tracks according to the obstacle distribution data and the candidate track information in the road passing area.

The obstacle distribution data may refer to position information of each static obstacle and each dynamic obstacle in a road passing area. The target trajectory may refer to a decision trajectory obtained after the trajectory decision.

Specifically, for each candidate trajectory, the trajectory cost value of the candidate trajectory, that is, the unreasonable degree of the candidate trajectory planning, may be determined based on candidate trajectory information of the candidate trajectory and obstacle distribution data in the road passing area. According to the track cost value of each candidate track, the candidate track with the minimum track cost value can be used as the target track.

Illustratively, each candidate trajectory in this embodiment, although different, may have the same sub-structure. For example, if two candidate trajectories are different only in the selection of the driving node on the k-th layer, when the quality of one of the candidate trajectories is evaluated, it is necessary to determine a cost value of a broken line segment 0 from the 0 th layer to the k-1 th layer and a cost value of a line segment a from the k-1 th layer to the k-th layer, and when the quality of the other trajectory is evaluated, it is necessary to determine a cost value of a broken line segment 0 from the 0 th layer to the k-1 th layer and a cost value of a line segment b from the k-1 th layer to the k-th layer, and it is thus apparent that the cost value of the broken line segment 0 is repeatedly calculated for a plurality of times, so that the case that the cost value of the line segment 0 is calculated for a plurality of times can be avoided based on a recursive manner, that is, in this embodiment, a Dynamic Programming (DP) manner can be based on obstacle distribution data in a road passing area and each candidate trajectoryAnd the track information is used for determining the target track of the target vehicle in running from all the tracks to be selected, so that the decision efficiency is improved under the condition of ensuring the accuracy. Wherein, the determination mode of dynamic programming DP can change the solving complexity from that required by adopting enumeration mode

Reduced to O (N)_layer·(N_par·N_node)²) And the efficiency of the track decision is greatly improved.

According to the technical scheme, the candidate track information of each candidate track in the road passing area of the structured road is determined, the candidate track information comprises the path information and the preset parameter information, namely the time information or the speed information, so that when the target track of the running target vehicle is determined according to the obstacle distribution data and the candidate track information in the road passing area, the track decision can be simultaneously carried out on the basis of the path and the speed or the time at one time, the more accurate, reasonable and feasible target track is obtained, and the decision accuracy is greatly improved.

On the basis of the above technical solution, S130 may include: determining at least one penalty parameter value corresponding to each to-be-selected track according to obstacle distribution data and each to-be-selected track information in a road passing area; determining a track cost value corresponding to each track to be selected according to each punishment parameter value; and determining the target track of the target vehicle according to the track cost values.

The penalty parameter value may be a value reflecting a degree of violation of a condition in the trajectory decision by the trajectory to be selected. The penalty parameter may be one or more, and the corresponding penalty parameter value may also be one or more. When only one penalty parameter value exists, the penalty parameter value can be directly used as a track cost value. The unreasonable degree of planning of each candidate track can be accurately evaluated through the track cost value, and the candidate track with the minimum track cost value is used as the target track, so that a more accurate target track can be obtained.

Example two

Fig. 3 is a flowchart of a trajectory decision method provided in the second embodiment of the present invention, and in this embodiment, optimization is performed on "determining at least one penalty parameter value corresponding to each trajectory to be selected according to obstacle distribution data and information of each trajectory to be selected in a road passing area" on the basis of the second embodiment. Wherein explanations of the same or corresponding terms as those of the above-described embodiments are omitted.

Referring to fig. 3, the trajectory decision method provided in this embodiment specifically includes the following steps:

s210, acquiring a road passing area of the structured road to be driven by the target vehicle.

S220, determining the to-be-selected track information of each to-be-selected track in the road passing area, wherein the to-be-selected track information comprises path information and preset parameter information, and the preset parameters comprise time information or speed information.

S230, determining a first punishment parameter value of each to-be-selected track in collision according to the obstacle distribution data in the road passing area and the path information in the to-be-selected track information.

The first penalty parameter value can be used as a safety index and used for representing the risk degree of collision between the target vehicle and the obstacle when the target vehicle runs along the track to be selected, namely the smaller the first penalty parameter value is, the smaller the risk degree of collision is.

Specifically, in the embodiment, the actual collision between the target vehicle and the obstacle and the situation that the target vehicle is close to the obstacle or enters the blind area although no collision occurs can be considered at the same time, so that the determined first penalty parameter value is more accurate and has a higher reference value.

Illustratively, S230 may include: determining a penalty value corresponding to each pre-divided space-time grid in the road passing area according to the obstacle distribution data in the road passing area, wherein the collision penalty value of the collision space-time grid where the obstacle in the road passing area is located is the largest; determining each target space-time grid covered by each trajectory to be selected according to the path information in the trajectory information to be selected; determining a target penalty value corresponding to each target space-time grid according to the penalty value corresponding to each space-time grid; and determining a first penalty parameter value of each to-be-selected track in collision according to each target penalty value.

Specifically, in order to uniformly consider two situations of actual collision and potential collision, the X-Y-T space corresponding to the road traffic area may be discretized into three-dimensional space-time grids. According to the position information and the speed information of each obstacle in the obstacle distribution data, determining a space-time grid occupied by each obstacle as a collision space-time grid, and presetting a collision penalty value C of each collision space-time grid_collision. The penalty value for other space-time grids near each collision space-time grid can be reduced along with the increase of the space-time distance between the collision space-time grids, for example, the penalty value is exponentially attenuated, so that the corresponding penalty value of each space-time grid can be obtained. For each candidate trajectory, each spatiotemporal grid covered by the candidate trajectory can be determined based on the path information of the candidate trajectory and used as a target spatiotemporal grid. And obtaining a target penalty value corresponding to each target space-time grid according to the corresponding relation between the space-time grids and the penalty values. According to the embodiment, all target penalty values can be added, and the addition result is used as a first penalty parameter value of the collision of the to-be-selected track; or averaging all target penalty values, and taking the obtained average value as a first penalty parameter value of the collision of the to-be-selected track. For example, the first penalty parameter value J for collision of the candidate trajectory can be determined by the following formula₁：

Wherein,

refers to the kth target penalty value; n is a radical of_sampleRefers to the number of targeted penalty values.

It should be noted that, when a three-dimensional space-time cost map is constructed, the penalty value C is caused by collision_collisionIs for punishing the actual collisionAnd the value of the first penalty parameter value is large enough to ensure that the first penalty parameter values of all the tracks not actually collided are smaller than the first penalty parameter value of any one collided track.

S240, determining a second punishment parameter value uncomfortable for each to-be-selected track according to the node coordinate of each driving node in the to-be-selected track information, the transverse offset of the driving node on each layer relative to the intersection point position and preset parameter information.

Wherein the intersection position may refer to a position where the boundary line of each layer intersects with the guiding line. The lateral offset is the distance between the travel node on each level and the intersection point location. The second penalty parameter value can be used as a comfort index for representing the discomfort degree of the target vehicle when the target vehicle runs along the candidate track, namely the discomfort degree is smaller when the second penalty parameter value is smaller. For example, the second penalty parameter value may be used to penalize the situation that the curvature of the candidate track is too large, and the lateral offset and the longitudinal speed of the candidate track frequently change, which may be specifically characterized by the curvature penalty parameter value, the offset penalty parameter value, and the speed penalty parameter value, respectively.

Illustratively, S240 may include the following steps S241-S244:

s241, according to the node coordinates of every three adjacent driving nodes in the to-be-selected track, determining the curvature of the driving nodes on each layer except the last layer in the to-be-selected track, and determining a curvature punishment parameter value according to each curvature.

The curvature penalty parameter value may be used to characterize a degree of an excessive curvature of the candidate trajectory, so as to penalty a case of the excessive curvature of the candidate trajectory.

In particular, a travel path constituted by travel nodes on each level

On the broken line locus, every three adjacent driving nodes Q_k-1-Q_k-Q_k+1It is possible to determine at the driving node Q_kCurvature k (Q) of_k)(k＝1,...,N_layer-1). The embodiment can passThe only circular arc of the three driving nodes is determined, and the reciprocal of the radius is the curvature. For example, according to Q_k-1、Q_kAnd Q_k+1The node coordinates of three driving nodes can determine Q_k-1Q_k、Q_kQ_k+1And Q_k-1Q_k+1Is the euclidean distance, and can be determined at the travel node Q by the following formula_kThe curvature of (a):

wherein a, b and c are each Q_k-1Q_k、Q_kQ_k+1And Q_k-1Q_k+1Length of (d). When the curvature at the driving node on each layer except the last layer is determined, all curvatures can be added, and the addition result is used as a curvature punishment parameter value; or averaging all curvatures, and using the obtained average value as a curvature penalty parameter value. For example, the curvature penalty parameter value J can be determined by the following formula_curvature：

And S242, determining an offset punishment parameter value according to the transverse offset of the driving node on each layer in the to-be-selected track information relative to the intersection point position.

The offset punishment parameter value can be used for representing the frequent change degree of the transverse offset of the to-be-selected track so as to punish the frequent change condition of the transverse offset of the to-be-selected track.

Specifically, the difference of the lateral offsets of two adjacent layers can be determined according to the lateral offset of the driving node on each layer relative to the intersection point position in the track information to be selected, and the offset punishment parameter value can be determined according to each difference. Illustratively, all the difference values may be added, and the addition result is taken as an offset penalty parameter value; or all the differences can be averaged to obtainAnd taking the obtained average value as an offset penalty parameter value. For example, the offset penalty parameter value J can be determined by the following equation_lateral：

Wherein, lat_kIs a driving node Q on the k-th floor_kA lateral offset from the intersection location; lat_k-1Is the driving node Q on the k-1 th floor_k-1A lateral offset with respect to the location of the intersection.

And S243, determining a speed punishment parameter value according to the preset parameter information of each driving node in the track information to be selected.

The speed penalty parameter value can be used for representing the degree of frequent change of the longitudinal speed of the trajectory to be selected, so as to penalty the frequent change of the longitudinal speed of the trajectory to be selected.

Specifically, when the preset parameter information is a time interval, the average longitudinal speed corresponding to each travel node may be determined based on the time interval. For example, the average longitudinal speed corresponding to the driving node may be determined by the following formula:

wherein,

is a driving node Q on the k-th floor_kAverage longitudinal velocity of (d); delta T_kIs a running node Q_kThe time interval of (c). And determining the difference value of the average longitudinal speeds of two adjacent running nodes according to the average longitudinal speed of each running node in each track to be selected, and determining a speed punishment parameter value according to each difference value. For example, all the difference values may be added, and the addition result is used as a speed penalty parameter value; or averaging all the differences, and taking the obtained average value as the speed penalty parameter value. For example, can be usedDetermining a speed penalty parameter value J by the following formula_longitudinal：

Wherein,

is a driving node Q on the k-th floor_kAverage longitudinal velocity of (d);

is the driving node Q on the k-1 th floor_k-1Average longitudinal velocity of (c). In the present embodiment, k may start from 2 or may start from 1. When k is 1, the above formula refers to

Is the average longitudinal speed at the starting position of the target vehicle, which needs to be determined based on the historical driving information. When k in the embodiment starts from 2, historical travel information does not need to be involved, so that the calculation is more convenient.

And S244, determining a second punishment parameter value which is uncomfortable for each track to be selected according to the curvature punishment parameter value, the offset punishment parameter value and the speed punishment parameter value.

Specifically, the uncomfortable second penalty parameter value J of each candidate trajectory can be determined through the following formula₂：

J₂＝w_curvature·J_curvature+w_lateral·J_lateral+w_longitudinal·J_longitudinal

Wherein, J_curvatureCurvature punishment parameter values of the to-be-selected track are obtained; j. the design is a square_lateralThe method comprises the steps of (1) punishing parameter values of offset of a to-be-selected track; j. the design is a square_longitudinalIs the speed punishment parameter value of the track to be selected; w is a_curvatureA curvature weight value that is a curvature penalty parameter value; w is a_lateralIs an offset weight value of the offset penalty parameter value;w_longitudinalis the velocity weight value of the velocity penalty parameter value. w is a_curvature、w_lateralAnd w_longitudinalAre all values greater than zero.

For each candidate trajectory, the operation of steps S241-S244 described above may be used to determine a second penalty parameter value that is uncomfortable for the candidate trajectory.

And S250, determining a third punishment parameter value of each to-be-selected track which does not reach the last layer according to preset parameter information in the to-be-selected track information.

And the third penalty parameter value can be used as an integrity index for representing the incomplete degree of the to-be-selected track, namely, the to-be-selected track which finally fails to reach the last layer is punished. The smaller the third penalty parameter value is, the smaller the incomplete degree of the to-be-selected track is.

In particular, there may be an infinite time interval Δ t or an average longitudinal velocity in the time dimension on a layer, such as the k-th layer

The case of 0 means that the target vehicle cannot reach the driving node on the next layer, i.e., the k +1 th layer within the allowable time, and therefore it is necessary to detect the first time that Δ t is infinite or the average longitudinal speed in the trajectory to be selected is the average longitudinal speed

And the stopping layer ordinal number is 0, and a corresponding third penalty parameter value can be determined according to the stopping layer ordinal number, namely, the larger the stopping layer ordinal number is, the closest the stopped position to the terminal point is indicated, and the smaller the third penalty parameter value is at the moment.

Illustratively, S250 may include: determining the first stop layer ordinal number with the average longitudinal speed of zero or infinite time interval in each track to be selected according to preset parameter information in the track information to be selected; and determining a third penalty parameter value of each to-be-selected track which does not reach the last layer according to the stop layer ordinal number, the preset layer number and the single-layer penalty value.

Specifically, for each candidate track, the selection method is based on the candidate track informationSequentially detecting whether the average longitudinal speed of each running node is zero or whether the time interval is infinite or not along the sequence from the starting point to the end point by using the preset parameter information of each running node; if yes, determining the sequence number of the running node as the stop layer sequence number N_stop(1≤N_stop≤N_layer). Determining that the target vehicle fails to go from N according to the stop layer ordinal number and the preset layer number_stopLayer integrity movement to Nth_layerAnd determining a third punishment parameter value of the to-be-selected track which does not reach the last layer according to the layer interval and a preset single-layer punishment value. For example, the third penalty parameter value J for the candidate trajectory not reaching the last layer can be determined by the following formula₃：

J₃＝(N_layer-N_stop)×C_stop

Wherein, C_stopIs a preset single-layer penalty value greater than zero.

And S260, determining the track cost value corresponding to each track to be selected according to the first penalty parameter value, the second penalty parameter value and the third penalty parameter value.

Specifically, the penalty parameter values corresponding to each candidate trajectory may be weighted and summed, and an obtained calculation result is used as the trajectory cost value corresponding to the candidate trajectory.

For example, the track cost value corresponding to each candidate track may be determined by the following formula:

J＝w₁·J₁+w₂·J₂+w₃·J₃

wherein J is the track cost value of the to-be-selected track; j. the design is a square₁The first penalty parameter value is a first penalty parameter value of the collision of the to-be-selected track; j. the design is a square₂The first penalty parameter value is uncomfortable for the trajectory to be selected; j. the design is a square₃The third punishment parameter value refers to that the trajectory to be selected does not reach the last layer; w is a₁A first weight value that is a first penalty parameter value; w is a₂A second weight value that is a second penalty parameter value; w is a₃Is a third weight value for a third penalty parameter value.

It should be noted that, the respective weight values used in the determination process of the trajectory cost value corresponding to each candidate trajectory, such as the curvature weight value w_curvatureOffset weight value w_lateralVelocity weight value w_longitudinalA first weight value w₁A second weight value w₂And a third weight value w₃And a penalty constant value for each non-weight, such as a collision penalty value C_collisionAnd a single-layer penalty value C_stopThe track decision method can be artificially set in advance based on business requirements and scenes, and can also be automatically determined in advance based on a Monte Carlo mode, so that the setting of penalty constant values of each weighted value and each non-weighted value is more accurate and reasonable, the determined track cost value is more accurate, the reference value is realized, and the accuracy of track decision is further improved.

And S270, determining the target track of the target vehicle according to the track cost values.

According to the technical scheme of the embodiment, the track cost value corresponding to each candidate track is measured by using the first punishment parameter value of each candidate track which is collided, the second punishment parameter value of each candidate track which is uncomfortable and the third punishment parameter value of each candidate track which does not reach the last layer, so that the advantages and disadvantages of the candidate tracks in the aspects of safety, comfort and integrity can be judged simultaneously, the determination of the target track is more reasonable and accurate, the optimal target track is convenient to obtain, and the accuracy of track decision is further improved.

EXAMPLE III

Fig. 4 is a flowchart of a trajectory decision method provided in the third embodiment of the present invention, and this embodiment describes in detail a determination process of "determining, based on a monte carlo manner, each weight value and each non-weight penalty constant value used in a determination process of a trajectory cost value corresponding to a trajectory to be selected". Wherein explanations of the same or corresponding terms as those of the above embodiments are omitted.

Referring to fig. 4, the trajectory decision method provided in this embodiment specifically includes the following steps:

s310, in the positive real number space, randomly selecting and combining the value of each weighted value and the value of each non-weighted penalty constant value used in the process of determining the track cost value corresponding to the to-be-selected track, and obtaining each to-be-selected combined result.

The positive real number space may be a set formed by real numbers greater than 0 preset based on service requirements. The weight value used in the process of determining the trajectory cost value corresponding to the candidate trajectory may be, but is not limited to, the curvature weight value w in embodiment two_curvatureOffset weight value w_lateralVelocity weight value w_longitudinalA first weight value w₁A second weight value w₂And a third weight value w₃. The non-weighted penalty constant value used in the process of determining the trajectory cost value corresponding to the trajectory to be selected may be, but is not limited to, the collision penalty value C in embodiment two_collisionAnd a single-layer penalty value C_stop. The candidate combination result may be a set composed of values of the respective weight values and values of the respective non-weight penalty constant values.

Specifically, the value of each weight value and the value of each non-weight penalty constant value may be randomly selected in the positive real number space based on a preset selection condition, and the values selected each time are combined into a to-be-selected combination result, so that all different to-be-selected combination results may be determined. The preset selection condition may include: the sum of the values of some weight values is 1, such as the first weight value w₁A second weight value w₂And a third weight value w₃The sum of (a) and (b) is 1.

It should be noted that, because potential conflicts may exist between some penalty parameter values, for example, the offset penalty parameter value corresponding to the to-be-selected track with a smaller curvature penalty parameter value is larger, a reasonable weight value needs to be set to perform discount and comprehensive consideration on various conflicts existing in the penalty parameter values, so that the determination of the track cost value is more accurate.

And S320, aiming at each candidate combination result, determining a corresponding sample target track under each candidate combination result from each sample candidate track according to the sample candidate track information and the sample obstacle distribution data of each sample candidate track of the target vehicle in the sample road passing area.

The sample road passing area may be a road passing area of the sample structured road, which is preset and used for determining penalty constant values of each weight value and each non-weight. In this embodiment, based on the preset hierarchical sampling manner provided in each embodiment, sample candidate trajectory information of each sample candidate trajectory in the sample road traffic area may be predetermined.

Specifically, for each weight value and each non-weight penalty constant value in each candidate combination result, based on a similar manner of target trajectory determination provided in the foregoing embodiments, a sample target trajectory may be determined from each sample candidate trajectory according to sample candidate trajectory information and sample obstacle distribution data of each sample candidate trajectory of the target vehicle in the sample road traffic area, so that an optimal sample candidate trajectory, that is, a sample target trajectory, obtained based on each candidate combination result may be determined.

S330, determining corresponding running errors according to the standard running track and each sample target track in the sample road running area.

The standard driving track may refer to a real track driven by a high-level driver in the sample road passing area.

Specifically, a sample area surrounded by each sample target track in the sample road passing area and a standard area surrounded by the standard running track in the sample road passing area can be determined, and a corresponding running error is obtained by comparing a difference value between each sample area and the standard area. When the running error is smaller, it is indicated that the sample target trajectory has a higher degree of similarity to the standard running trajectory, i.e., the driving level of the target vehicle is closer to a driver of a high level.

And S340, determining a target combination result from each to-be-selected combination result according to each driving error, and obtaining each weight value and each non-weight penalty constant value used in the determination process of the track cost value corresponding to the to-be-selected track according to the target combination result.

Specifically, the to-be-selected combination result corresponding to the sample target trajectory with the smallest running error may be determined as the target combination result, and specific values of the penalty constant values of each weight value and each non-weight value are obtained based on the target combination result, so that a reasonable value is automatically determined.

And S350, acquiring a road passing area of the structured road to be driven by the target vehicle.

And S360, determining the to-be-selected track information of each to-be-selected track in the road passing area, wherein the to-be-selected track information comprises path information and preset parameter information, and the preset parameters comprise time information or speed information.

S370, determining at least one penalty parameter value corresponding to each to-be-selected track according to the obstacle distribution data and the to-be-selected track information in the road passing area.

Specifically, the operations of steps S310-S340 may be utilized to automatically obtain respective weight values, such as curvature weight value w, used in determining penalty parameter values in step S370_curvatureOffset weight value w_lateralVelocity weight value w_longitudinalAnd a penalty constant value for each non-weight, such as a collision penalty value C_collisionAnd a single-layer penalty value C_stopAnd determining at least one penalty parameter value corresponding to each trajectory to be selected based on the weight values and the penalty constant values.

And S380, determining the track cost value corresponding to each track to be selected according to each punishment parameter value.

Specifically, the operations of steps S310-S340 may be utilized to determine the weight values used in determining the track cost value based on the penalty parameter values in step S380, such as: first weight value w₁A second weight value w₂And a third weight value w₃And determining the track cost value corresponding to each track to be selected based on the weight values.

It should be noted that, values of penalty constant values of each weight value and each non-weight selected for different types of vehicles are different, so that each weight value and each non-weight penalty constant value corresponding to each vehicle can be determined in advance, and a corresponding relationship between a type identifier of each vehicle and each weight value and each non-weight penalty constant value is established, so that specific values of each weight value and each non-weight penalty constant value corresponding to a target vehicle can be obtained more quickly in the following based on the type identifier of the target vehicle and the corresponding relationship.

And S390, determining the target track of the target vehicle according to the track cost values.

According to the technical scheme of the embodiment, all the conditions of the combination results to be selected are determined in the positive real number space, and an optimal target combination result is selected from all the combination results to be selected based on the standard running track in the sample road traffic area, so that the penalty constant values of all the weighted values and all the non-weighted values can be set reasonably and accurately, and the accuracy of track decision is improved.

On the basis of the above technical solution, S320 may include: for each candidate combination result, determining a sample cost value corresponding to each sample candidate track according to sample candidate track information and sample obstacle distribution data of each sample candidate track of the target vehicle in the sample road traffic area; screening out the combination results to be selected which meet the preset punishment condition from the combination results to be selected according to the cost value of each sample, and taking the selected combination results as candidate combination results; and determining a corresponding sample target track under each candidate combination result according to the sample cost value corresponding to each sample candidate track determined according to each candidate combination result.

The preset penalty condition may be a penalty logic that is preset to be satisfied with the trajectory to be selected based on the service scenario, so that the setting of the penalty constant values of each weight value and each non-weight value is more reasonable. Illustratively, the predetermined penalty condition may include, but is not limited to: the sample cost value corresponding to any one sample candidate track which does not collide and reaches the last layer is less than that corresponding to any sample candidate track which collides or does not reach the last layer; the sample cost value corresponding to any sample candidate track without collision is smaller than the sample cost value corresponding to any sample candidate track with collision; and the sample cost value corresponding to any sample candidate track reaching the last layer is smaller than the sample cost value corresponding to any sample candidate track not reaching the last layer.

In particular, a penalty constant value, such as a collision penalty value C, is set for each non-weight_collisionAnd a single-layer penalty value C_stopTime, impact penalty value C_collisionAnd a single-layer penalty value C_stopAll require values as large as possible, but due to collision penalty value C_collisionAnd a single-layer penalty value C_stopThe two parameters are unrelated and have different action modes, so that the sizes of the two parameters cannot be directly commented, whether the specific values of the two parameters meet reasonable punishment logic or not can be detected based on a preset punishment condition, and the candidate combination results meeting the preset punishment condition are used as candidate combination results, so that the candidate combination results meeting the preset punishment condition are firstly screened out from the candidate combination results, and then the corresponding sample target track is determined according to each candidate combination result, so that the punishment constant value is set more reasonably.

Accordingly, the step S340 of determining the target combination result from the candidate combination results according to the driving errors may include: and determining a target combination result from the candidate combination results according to the running errors. By aiming at each candidate combination result meeting the preset punishment condition, the corresponding running error is determined according to the standard running track in the sample road passing area and the sample target track corresponding to each candidate combination result, and the candidate combination result corresponding to the sample target track with the minimum running error can be determined as the target combination result, so that the optimal target combination result can be determined from each candidate combination result meeting the preset punishment condition, the rationality and the accuracy of the target combination result are further improved, and the accuracy of the track decision is further improved.

The following is an embodiment of a trajectory decision device provided in an embodiment of the present invention, which belongs to the same inventive concept as the trajectory decision method in the foregoing embodiments, and details that are not described in detail in the embodiment of the trajectory decision device may refer to the embodiment of the trajectory decision method.

Example four

Fig. 5 is a schematic structural diagram of a trajectory decision device according to a fourth embodiment of the present invention, where the present embodiment is applicable to a situation of performing trajectory decision on an automatically driven vehicle on a structured road, and the device specifically includes: a road passing area obtaining module 410, a candidate track information determining module 420 and a target track determining module 430.

The road passing area acquiring module 410 is configured to acquire a road passing area of a structured road where a target vehicle is to travel; a candidate trajectory information determining module 420, configured to determine candidate trajectory information of each candidate trajectory in a road traffic area, where the candidate trajectory information includes path information and preset parameter information, and the preset parameter includes time information or speed information; and the target track determining module 430 is configured to determine a target track where the target vehicle runs from the candidate tracks according to the obstacle distribution data and the candidate track information in the road passing area.

Optionally, the candidate trajectory information determining module 420 is specifically configured to: according to the preset number of layers, layering the guiding lines in the road traffic area, and taking the normal line at the intersection point position of each layer and the guiding line as the boundary line of the corresponding layer; dividing the boundary of each layer according to the number of preset nodes, and taking each divided node as a running node of the target vehicle on the boundary; determining each parameter value to be selected when the target vehicle runs to a running node according to the preset parameter value quantity; and respectively selecting a driving node on the boundary line of each layer, respectively selecting a parameter value to be selected for each selected driving node, and taking each selected driving node and each parameter value to be selected as the track information to be selected of a track to be selected.

Optionally, the target trajectory determination module 430 includes:

the penalty parameter value determining unit is used for determining at least one penalty parameter value corresponding to each to-be-selected track according to the barrier distribution data and the to-be-selected track information in the road passing area;

the track cost value determining unit is used for determining the track cost value corresponding to each track to be selected according to each punishment parameter value;

and the target track determining unit is used for determining the target track of the running target vehicle according to the track cost values.

Optionally, the penalty parameter value determining unit includes:

the first penalty parameter value determining subunit is used for determining a first penalty parameter value of each to-be-selected track in collision according to the obstacle distribution data in the road passing area and the path information in the to-be-selected track information;

a second punishment parameter value determining subunit, configured to determine, according to the node coordinate of each driving node in the to-be-selected track information, the lateral offset of the driving node on each layer with respect to the intersection position, and preset parameter information, a second punishment parameter value that is uncomfortable for each to-be-selected track;

and the third punishment parameter value determining subunit is used for determining the third punishment parameter value of each to-be-selected track which does not reach the last layer according to the preset parameter information in the to-be-selected track information.

Optionally, the trajectory cost value determining unit is specifically configured to: determining the track cost value corresponding to each candidate track through the following formula:

J＝w₁·J₁+w₂·J₂+w₃·J₃

Optionally, the first penalty parameter value determining subunit is specifically configured to: determining a penalty value corresponding to each pre-divided space-time grid in the road passing area according to the obstacle distribution data in the road passing area, wherein the collision penalty value of the collision space-time grid where the obstacle in the road passing area is located is the largest; determining each target space-time grid covered by each trajectory to be selected according to the path information in the trajectory information to be selected; determining a target penalty value corresponding to each target space-time grid according to the penalty value corresponding to each space-time grid; and determining a first penalty parameter value of each to-be-selected track in collision according to each target penalty value.

Optionally, the second penalty parameter value determining subunit is specifically configured to: determining the curvature of the driving node on each layer except the last layer in the trajectory to be selected according to the node coordinates of every three adjacent driving nodes in the trajectory to be selected, and determining a curvature punishment parameter value according to each curvature; determining an offset punishment parameter value according to the transverse offset of the driving node on each layer in the to-be-selected track information relative to the intersection point position; determining a speed punishment parameter value according to preset parameter information of each driving node in the track information to be selected; and determining a second penalty parameter value which is uncomfortable for each track to be selected according to the curvature penalty parameter value, the offset penalty parameter value and the speed penalty parameter value.

Optionally, the third penalty parameter value determining subunit is specifically configured to: determining the first stop layer ordinal number with the average longitudinal speed of zero or infinite time interval in each track to be selected according to preset parameter information in the track information to be selected; and determining a third penalty parameter value of each to-be-selected track which does not reach the last layer according to the stop layer ordinal number, the preset layer number and the single-layer penalty value.

Optionally, the apparatus further comprises:

and the numerical value determining module is used for determining each weight value and each non-weight penalty constant value used in the determination process of the track cost value corresponding to the to-be-selected track based on a Monte Carlo mode before determining the target track driven by the target vehicle from each to-be-selected track according to the obstacle distribution data and each to-be-selected track information in the road passing area.

Optionally, the value determining module includes:

a candidate combination result determining unit, configured to randomly select and combine a value of each weight value and a value of each unweighted penalty constant value used in the determination process of the trajectory cost value corresponding to the candidate trajectory in the positive real number space, and obtain each candidate combination result;

the sample target track determining unit is used for determining a corresponding sample target track under each to-be-selected combination result from each sample to-be-selected track according to sample to-be-selected track information and sample obstacle distribution data of each sample to-be-selected track of the target vehicle in the sample road passing area;

the driving error determining unit is used for determining corresponding driving errors according to the standard driving track and each sample target track in the sample road passing area;

and the numerical value determining unit is used for determining a target combination result from each to-be-selected combination result according to each driving error, and obtaining each weight value and each non-weight punishment constant value used in the determination process of the track cost value corresponding to the to-be-selected track according to the target combination result.

Optionally, the sample target trajectory determining unit is specifically configured to: for each candidate combination result, determining a sample cost value corresponding to each sample candidate track according to sample candidate track information and sample obstacle distribution data of each sample candidate track of the target vehicle in the sample road traffic area; screening out the combination results to be selected which meet the preset punishment condition from the combination results to be selected according to the cost value of each sample, and taking the selected combination results as candidate combination results; determining a sample target track corresponding to each candidate combination result according to the sample cost value corresponding to each sample candidate track determined according to each candidate combination result; accordingly, the numerical value determination unit is specifically configured to: and determining a target combination result from the candidate combination results according to the running errors.

Optionally, the penalty condition is preset, and includes:

the sample cost value corresponding to any one sample candidate track which does not collide and reaches the last layer is less than that corresponding to any sample candidate track which collides or does not reach the last layer; the sample cost value corresponding to any sample candidate track without collision is smaller than the sample cost value corresponding to any sample candidate track with collision; and the sample cost value corresponding to any sample candidate track reaching the last layer is smaller than the sample cost value corresponding to any sample candidate track not reaching the last layer.

The trajectory decision device provided by the embodiment of the invention can execute the trajectory decision method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects for executing the trajectory decision method.

EXAMPLE five

Fig. 6 is a schematic structural diagram of an apparatus according to a fifth embodiment of the present invention. Fig. 6 illustrates a block diagram of an exemplary device 12 suitable for use in implementing embodiments of the present invention. The device 12 shown in fig. 6 is only an example and should not bring any limitations to the functionality and scope of use of the embodiments of the present invention.

As shown in FIG. 6, device 12 is in the form of a general purpose computing device. The components of device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including the system memory 28 and the processing unit 16.

Bus 18 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, such architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, micro-channel architecture (MAC) bus, enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus.

Device 12 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by device 12 and includes both volatile and nonvolatile media, removable and non-removable media.

The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)30 and/or cache memory 32. Device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 6, and commonly referred to as a "hard drive"). Although not shown in FIG. 6, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. System memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.

A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 42 generally carry out the functions and/or methodologies of the described embodiments of the invention.

Device 12 may also communicate with one or more external devices 14 (e.g., keyboard, pointing device, display 24, etc.), with one or more devices that enable a user to interact with device 12, and/or with any devices (e.g., network card, modem, etc.) that enable device 12 to communicate with one or more other computing devices. Such communication may be through an input/output (I/O) interface 22. Also, the device 12 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via the network adapter 20. As shown, the network adapter 20 communicates with the other modules of the device 12 via the bus 18. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

The processing unit 16 executes various functional applications and data processing by running a program stored in the system memory 28, for example, to implement a trajectory decision method provided by the embodiment of the present invention, the method includes:

determining candidate track information of each candidate track in a road passing area, wherein the candidate track information comprises path information and preset parameter information, and the preset parameter comprises time information or speed information;

and determining a target track of the running of the target vehicle from the candidate tracks according to the obstacle distribution data and the candidate track information in the road passing area.

Of course, those skilled in the art can understand that the processor can also implement the technical solution of the trajectory decision method provided by any embodiment of the present invention.

EXAMPLE six

A sixth embodiment provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the trajectory decision method steps provided in any embodiment of the present invention, and the method includes:

Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer-readable storage medium may be, for example but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, RF, etc., or any suitable combination of the foregoing.

Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

It will be understood by those skilled in the art that the modules or steps of the invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device or distributed across a network of computing devices, and optionally they may be implemented by program code executable by a computing device, such that it may be stored in a memory device and executed by a computing device, or it may be separately fabricated into various integrated circuit modules, or it may be fabricated by fabricating a plurality of modules or steps thereof into a single integrated circuit module. Thus, the present invention is not limited to any specific combination of hardware and software.

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A trajectory decision method, comprising:

2. The method of claim 1, wherein determining candidate trajectory information for each candidate trajectory in the road passing area comprises:

layering the guiding lines in the road traffic area according to a preset number of layers, and taking a normal line at the intersection point position of each layer and the guiding lines as a boundary line of the corresponding layer;

dividing a boundary of each layer according to the number of preset nodes, and taking each divided node as a running node of the target vehicle on the boundary;

determining each parameter value to be selected when the target vehicle runs to the running node according to the preset parameter value quantity;

and respectively selecting a driving node on the boundary line of each layer, respectively selecting a parameter value to be selected for each selected driving node, and taking each selected driving node and each parameter value to be selected as the track information to be selected of a track to be selected.

3. The method according to claim 1, wherein determining a target trajectory traveled by the target vehicle from among the candidate trajectories based on obstacle distribution data in the road passing area and the candidate trajectory information includes:

determining at least one penalty parameter value corresponding to each to-be-selected track according to obstacle distribution data and each to-be-selected track information in the road passing area;

determining a track cost value corresponding to each to-be-selected track according to each punishment parameter value;

and determining a target track of the target vehicle according to the track cost values.

4. The method according to claim 3, wherein determining at least one penalty parameter value corresponding to each candidate trajectory according to obstacle distribution data in the road passing area and information of each candidate trajectory comprises:

determining a first punishment parameter value of each to-be-selected track in collision according to obstacle distribution data in the road passing area and path information in the to-be-selected track information;

determining a second punishment parameter value uncomfortable for each to-be-selected track according to the node coordinate of each driving node in the to-be-selected track information, the transverse offset of the driving node on each layer relative to the intersection point position and preset parameter information;

and determining a third punishment parameter value of each to-be-selected track which does not reach the last layer according to preset parameter information in the to-be-selected track information.

5. The method according to claim 4, wherein the track cost value corresponding to each candidate track is determined by the following formula:

J＝w₁·J₁+w₂·J₂+w₃·J₃

wherein J refers to the track cost value of the to-be-selected track; j. the design is a square₁The first penalty parameter value refers to the collision of the trajectory to be selected; j. the design is a square₂The second penalty parameter value is uncomfortable for the trajectory to be selected; j. the design is a square₃The third penalty parameter value refers to that the trajectory to be selected does not reach the last layer; w is a₁Is a first weight of the first penalty parameter valueA weight value; w is a₂A second weight value that is the second penalty parameter value; w is a₃Is a third weight value for the third penalty parameter value.

6. The method according to claim 4, wherein determining a first penalty parameter value for each candidate trajectory in collision according to obstacle distribution data in the road passing area and path information in the candidate trajectory information comprises:

determining a penalty value corresponding to each pre-divided space-time grid in the road passing area according to the obstacle distribution data in the road passing area, wherein the collision penalty value of the collision space-time grid where the obstacle in the road passing area is located is the largest;

determining each target space-time grid covered by each trajectory to be selected according to path information in the trajectory information to be selected;

determining a target penalty value corresponding to each target space-time grid according to the penalty value corresponding to each space-time grid;

and determining a first penalty parameter value of each to-be-selected track in collision according to each target penalty value.

7. The method according to claim 4, wherein determining a second penalty parameter value which is uncomfortable for each candidate trajectory according to the node coordinates of each driving node in the candidate trajectory information, the lateral offset of the driving node on each layer relative to the intersection position and preset parameter information comprises:

determining curvatures of driving nodes on each layer except the last layer in the to-be-selected track according to the node coordinates of every three adjacent driving nodes in the to-be-selected track, and determining curvature punishment parameter values according to the curvatures;

determining an offset punishment parameter value according to the transverse offset of the driving node on each layer in the to-be-selected track information relative to the intersection point position;

determining a speed punishment parameter value according to preset parameter information of each driving node in the to-be-selected track information;

and determining a second penalty parameter value which is uncomfortable for each track to be selected according to the curvature penalty parameter value, the offset penalty parameter value and the speed penalty parameter value.

8. The method of claim 4, wherein determining a third penalty parameter value for each candidate track not reaching the last layer according to preset parameter information in the candidate track information comprises:

determining the first stop layer ordinal number with zero average longitudinal speed or infinite time interval in each to-be-selected track according to preset parameter information in the to-be-selected track information;

and determining a third penalty parameter value of each to-be-selected track which does not reach the last layer according to the stop layer ordinal number, the preset layer number and the single-layer penalty value.

9. The method according to claim 3, before determining a target trajectory traveled by the target vehicle from among the candidate trajectories based on obstacle distribution data in the road passing area and the candidate trajectory information, further comprising:

and determining each weight value and each non-weight penalty constant value used in the determination process of the track cost value corresponding to the to-be-selected track based on a Monte Carlo mode.

10. The method according to claim 9, wherein determining penalty constant values of each weight value and each non-weight used in the determination process of the trajectory cost value corresponding to the trajectory to be selected based on a monte carlo manner includes:

randomly selecting and combining values of each weight value and each non-weight penalty constant value used in the process of determining the track cost value corresponding to the to-be-selected track in a positive real number space to obtain each to-be-selected combination result;

for each to-be-selected combination result, determining a corresponding sample target track under each to-be-selected combination result from each sample to-be-selected track according to sample to-be-selected track information and sample obstacle distribution data of each sample to-be-selected track of the target vehicle in a sample road passing area;

determining corresponding running errors according to the standard running tracks in the sample road passing area and each sample target track;

and determining a target combination result from each to-be-selected combination result according to each driving error, and obtaining each weight value and each non-weight penalty constant value used in the determination process of the track cost value corresponding to the to-be-selected track according to the target combination result.

11. The method according to claim 10, wherein for each candidate combination result, determining a sample target trajectory corresponding to each candidate combination result from each sample candidate trajectory according to sample candidate trajectory information and sample obstacle distribution data of each sample candidate trajectory of the target vehicle in a sample road passing area, includes:

for each to-be-selected combination result, determining a sample cost value corresponding to each sample to-be-selected track according to sample to-be-selected track information and sample obstacle distribution data of each sample to-be-selected track of the target vehicle in a sample road passing area;

screening out the combination results to be selected which meet a preset punishment condition from the combination results to be selected according to the cost value of each sample, and taking the selected combination results as candidate combination results;

determining a sample target track corresponding to each candidate combination result according to the sample cost value corresponding to each sample candidate track determined according to each candidate combination result;

correspondingly, according to each driving error, determining a target combination result from each to-be-selected combination result, which comprises the following steps:

and determining a target combination result from the candidate combination results according to the running errors.

12. The method of claim 11, wherein the predetermined penalty condition comprises:

the sample cost value corresponding to any one sample candidate track which does not collide and reaches the last layer is less than that corresponding to any sample candidate track which collides or does not reach the last layer;

the sample cost value corresponding to any sample candidate track without collision is smaller than the sample cost value corresponding to any sample candidate track with collision;

and the sample cost value corresponding to any sample candidate track reaching the last layer is smaller than the sample cost value corresponding to any sample candidate track not reaching the last layer.

13. A trajectory decision device, comprising:

14. An apparatus, characterized in that the apparatus comprises:

one or more processors;

a memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement the trajectory decision method of any one of claims 1-12.

15. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a trajectory decision method according to any one of claims 1-12.