JP7703928B2 - Travel trajectory generation method, travel assistance method, travel trajectory generation device, and travel assistance device - Google Patents

Description

The present invention relates to a driving trajectory generation method, a driving support method, a driving trajectory generation device, and a driving support device.

Patent Document 1 below proposes a vehicle driving control device that stores the driving trajectory when the vehicle is driven manually and sets the driving trajectory as the target trajectory of the vehicle. The vehicle driving control device recognizes the lane in which the vehicle is driving, and when the vehicle can maintain the lane while driving, causes the vehicle to maintain the lane while driving along the target trajectory.

JP 2018-22353 A

However, the vehicle driving control device described in Patent Document 1 may not be able to set an appropriate target driving trajectory. For example, when the vehicle cannot recognize the lane in which the vehicle is traveling, the vehicle sets the target driving trajectory based on the driving path in the past, which may not be an optimal trajectory.
The present invention aims to set an optimized target driving trajectory based on the driving path of a vehicle traveled on the same road in the past.

In one aspect of the present invention, a driving trajectory generation method obtains driving trajectories when a vehicle travels a specified section of the same road multiple times, extracts multiple passing points that are positions on the driving trajectory for each of the obtained driving trajectories, assigns a reliability to each of the multiple passing points, associates the passing points with their reliability and stores them in a storage device, and generates multiple driving trajectory candidates for traveling the specified section by connecting the passing points stored in the storage device. Of the multiple generated driving trajectory candidates, one of the driving trajectory candidates is extracted based on the sum of the reliability of the passing points belonging to the driving trajectory candidate, and generates the extracted driving trajectory candidate as a target driving trajectory along which the vehicle will travel.

According to the present invention, an optimized target driving trajectory can be set based on the driving trajectory when the vehicle traveled on the same road in the past.

1 is a diagram showing an example of a schematic configuration of a vehicle equipped with a driving support device according to an embodiment; 2 is a block diagram of an example of a functional configuration of the driving control device of FIG. 1 . FIG. 2 is an explanatory diagram of an example of map information. 1 is an explanatory diagram of links representing going straight, turning right, and turning left in an intersection area. FIG. FIG. 11 is an explanatory diagram of an example of a method for determining the connectivity between nodes on a travel trajectory; FIG. 2 is an explanatory diagram of an example of a travel trajectory generation method according to an embodiment; FIG. 11 is an explanatory diagram of an example of a method for generating a travel trajectory candidate. FIG. 11 is an explanatory diagram of an example of a method for generating a travel trajectory candidate. FIG. 11 is an explanatory diagram of an example of a method for determining independence between travel trajectory candidates; FIG. 11 is an explanatory diagram of an example of a method for determining independence between travel trajectory candidates;

Below, an embodiment of the present invention will be described with reference to the drawings. In the description of the drawings, identical or similar parts are given the same or similar reference numerals, and duplicate explanations will be omitted. Each drawing is schematic and may differ from the actual product. The embodiments shown below are examples of devices and methods for embodying the technical idea of the present invention, and the technical idea of the present invention is not limited to the devices and methods exemplified in the following embodiments. The technical idea of the present invention can be modified in various ways within the technical scope described in the claims.

(composition)
1 shows an example of a schematic configuration of a vehicle equipped with a driving assistance device according to an embodiment. The driving assistance device assists the driving of the host vehicle 1 along a target driving trajectory. The driving assistance control of the host vehicle 1 by the driving assistance device according to the embodiment may include, for example, autonomous driving control that automatically drives the host vehicle 1 along a target driving trajectory without the involvement of an occupant (e.g., a driver). In addition, the driving assistance control by the driving assistance device may include, for example, driving assistance control that controls at least the steering angle of the host vehicle 1 to assist the host vehicle 1 in driving along the target driving trajectory.

The driving assistance device of the embodiment includes a vehicle speed sensor 2, a yaw rate sensor 3, a vehicle position acquisition device 4, a lane recognition device 5, an actuator 6, and a driving control device 7.
The vehicle speed sensor 2 is a vehicle sensor (a sensor that detects the running state of the host vehicle 1) that detects the wheel speed of the host vehicle 1 and calculates the vehicle speed of the host vehicle 1 based on the wheel speed.
The yaw rate sensor 3 is a vehicle sensor that detects the yaw rate generated in the host vehicle 1 .

The vehicle position acquisition device 4 measures the current position of the vehicle 1. The vehicle position acquisition device 4 may include, for example, a Global Navigation System (GNSS) receiver. The GNSS receiver is, for example, a Global Positioning System (GPS) receiver, and receives radio waves from a plurality of navigation satellites to measure the current position of the vehicle 1. The vehicle position acquisition device 4 may be, for example, an inertial navigation system.
The lane recognition device 5 detects lane boundary lines (e.g., line marks such as white lines or yellow lines) of the lane in which the vehicle 1 is traveling using object sensors such as a camera, laser radar, and LiDAR (Light Detection and Ranging) that detect objects around the vehicle, and recognizes the lane in which the vehicle 1 is traveling.

The actuator 6 generates vehicle behavior of the host vehicle 1 by operating the steering device, drive device, and braking device of the host vehicle 1 in response to a control signal from the cruise control device 7. The actuator 6 includes a steering actuator, an accelerator opening actuator, and a brake control actuator.
The steering actuator operates a steering device to control the steering direction and steering amount of the vehicle. The accelerator opening actuator operates a drive device such as an engine and a drive motor to control the acceleration of the host vehicle 1. The brake control actuator operates a braking device to control the deceleration of the host vehicle 1.

The driving control device 7 is an electronic control unit (ECU) that performs driving assistance control of the host vehicle 1. For example, when autonomous driving control is performed as driving assistance control, the driving control device 7 automatically controls the driving of the host vehicle 1 by driving the actuator 6 based on the driving environment around the host vehicle 1.
Further, for example, when driving assistance control is executed as driving assistance control, the driving control device 7 controls at least the steering angle of the host vehicle 1 by driving the actuator 6 based on the driving environment around the host vehicle 1 .
The driving control device 7 includes a processor 10 and peripheral components such as a storage device 11. The processor 10 may be, for example, a central processing unit (CPU) or a micro-processing unit (MPU). The storage device 11 may include a semiconductor storage device, a magnetic storage device, an optical storage device, or the like.
The functions of the driving control device 7 described below are realized, for example, by the processor 10 executing a computer program stored in the storage device 11. Note that the driving control device 7 may be formed by dedicated hardware for executing each of the information processes described below.

2 is a block diagram of an example of a functional configuration of the driving control device 7. The driving control device 7 includes a driving trajectory storage unit 20, a trajectory optimization unit 21, a reliability calculation unit 22, a map information storage unit 23, a link assignment unit 24, a combination optimization unit 25, a trajectory candidate storage unit 26, a driving road determination unit 27, a road shape estimation unit 28, and a vehicle control unit 29.
The cruise control device 7 may operate, for example, in at least the following three operating modes:
(1) A driving trajectory acquisition mode in which a driving trajectory of the vehicle 1 is accumulated during manual driving (2) A trajectory candidate generation mode in which the accumulated driving trajectory is optimized to generate a driving trajectory candidate (3) A driving assistance mode in which driving of the vehicle 1 along the driving trajectory candidate is supported Note that, for example, the processing in the trajectory candidate generation mode may be performed by offline processing. The processing by the driving control device 7 in the operation modes will be described below.

(Trajectory acquisition mode)
In the travel trajectory acquisition mode, while the host vehicle 1 is being driven manually, the travel control device 7 acquires the positions of the left and right lane boundary lines of the lane in which the host vehicle 1 is traveling from the lane recognition device 5. The travel control device 7 also acquires the vehicle speed and yaw rate of the host vehicle 1 detected by the vehicle speed sensor 2 and yaw rate sensor 3 via a CAN (Controller Area Network) 8. The travel control device 7 also acquires position information of the host vehicle 1 from the host vehicle position acquisition device 4. The travel control device 7 accumulates time-series data in which combinations of these pieces of data are associated with time information at the time of acquisition in the travel trajectory storage unit 20.

(Trajectory candidate generation mode)
After a sufficient amount of travel trajectories have been accumulated in the travel trajectory acquisition mode, the travel control device 7 optimizes the travel trajectory through offline processing in the trajectory candidate generation mode to generate a travel trajectory candidate.
The travel trajectory stored in the travel trajectory acquisition mode is an estimation result of the vehicle position acquisition device 4 performing online processing to estimate the position of the vehicle 1, and the position information may contain errors. For example, in a location where the reception condition of the GNSS receiver is poor, an error may occur in the position information. For this reason, the trajectory optimization unit 21 reads the travel trajectory from the travel trajectory storage unit 20, and optimizes the shape of the travel trajectory by performing bundle adjustment that solves an optimization problem to minimize errors in the position information based on the position information acquired by the vehicle position acquisition device 4 and the vehicle speed and yaw rate acquired from the CAN 8.

The trajectory optimization unit 21 samples points (x, y) on the optimized driving trajectory and the attitude θ of the vehicle 1 at those points at a predetermined interval (e.g., 1 meter). Hereinafter, the sampled points on the driving trajectory (i.e., passing points, driving points) are referred to as "nodes." The attitude θ of the vehicle 1 at each node may be, for example, the traveling direction of the vehicle 1 at that node, or the vehicle attitude angle.

The reliability calculation unit 22 calculates the center position of the lane in which the vehicle 1 is traveling from the recognition result of the lane boundary line that is linked to the position information of the traveling trajectory by the time information and stored in the traveling trajectory storage unit 20. For example, the midpoint of the positions of the left and right lane boundary lines expressed in the vehicle coordinate system may be calculated as the center position of the lane.
In the vehicle coordinate system, the position of the vehicle 1 is the origin of the coordinate system, so the distance between the midpoint of the positions of the left and right lane boundary lines and the origin is the lateral deviation of the node from the center of the lane.

The reliability calculation unit 22 calculates a higher reliability (score) as the lateral deviation of the node from the center of the road lane is smaller.
For example, the reliability calculation unit 22 determines whether or not the host vehicle 1 is traveling approximately in the center of the lane at each node based on the lateral deviation.
Here, since a wide lane generally allows for some deviation in the vehicle's driving line compared to a narrow lane, it may be determined whether the vehicle 1 is located approximately in the center or not depending on the lane width. For example, if the detected lane width (the distance between the left and right white lines) is W (m) and the general width of a vehicle is 1.8 m, the threshold value may be set to max((W-1.8)/2-0.4, 0.2), and if the lateral deviation is equal to or less than this threshold value, it may be determined that the vehicle 1 is driving approximately in the center of the lane.
If it is determined that the vehicle 1 is traveling approximately in the center of the lane, a reliability of +1 is given to this node, and if it is determined that the vehicle 1 is not traveling approximately in the center of the lane, a reliability of -1 is given. Also, if the lane has not been recognized, the position of the lane cannot be known, so a reliability of 0 is given.
Through the above processing, each node has information on the position and orientation (x, y, θ) and reliability {-1, 0, +1}. Note that the reliability may be continuously increased as the lateral deviation of the node from the center of the lane of the road becomes smaller, or may be increased in stages. In other words, the reliability may be calculated so that it is higher when the lateral deviation of the node from the center of the lane of the road is small than when it is large.

Map information having position information for each road is stored in the map information storage unit 23. The map information stored in the map information storage unit 23 may be, for example, map information for navigation.
Fig. 3 is an explanatory diagram of an example of map information stored in the map information storage unit 23. The map information stored in the map information storage unit 23 has a graph structure made up of map nodes Nm (represented by black plots in Figs. 3 and 4) and map links Lm (represented by arrows in Figs. 3 and 4) connecting the map nodes.

The intersection areas Ac1, Ac2 have different road classifications from other areas (single roads), and the graph structure within the intersection areas Ac1, Ac2 is divided into map links that go straight through the intersection areas and map links that turn right and/or left, and each of these map links is represented by multiple map links.
4, for example, in a crossroads intersection area Ac1, the map links going straight through the intersection area are represented by three map links Lm1, Lm2, and Lm3. The map links turning right are represented by three map links Lm1, Lm4, and Lm5 including the diagonal map link Lm4. The map links turning left are represented by two map links Lm1 and Lm6.

In this embodiment, in order to facilitate the processing described below, a plurality of map links from the entrance to the exit of the intersection areas Ac1 and Ac2 are treated as a single map link.
For example, the three map links Lm1, Lm2, and Lm3 are connected and treated as a single map link LS that goes straight ahead, the three map links Lm1, Lm4, and Lm5 are connected and treated as a single map link LR that turns right, and the two map links Lm1 and Lm6 are connected and treated as a single map link LL that turns left.
By treating multiple map links connected within an intersection as a single map link in this manner, the accuracy of assigning travel trajectories to map links is improved, as will be described later.

Please refer to Fig. 2. The link assignment unit 24 assigns a node sampled from the travel path to the map link closest to this node. The link assignment unit 24 divides the travel path at the node where the assigned map link changes. In this way, the travel path is divided into predetermined sections corresponding to a single map link.
3. The dashed line indicates a travel trajectory that enters the intersection area Ac1 from the bottom of the drawing, turns right, and then turns left at the intersection area Ac2 and heads toward the top of the drawing. The link assignment unit 24 divides this travel trajectory into a travel trajectory S1 that travels on a section of a single road that enters the intersection area Ac1 from the bottom of the drawing, a travel trajectory S2 that turns right within the intersection area Ac1, a travel trajectory S3 that travels on a section of a single road that exits the intersection area Ac1 and enters the intersection area Ac2, a travel trajectory S4 that turns left within the intersection area Ac2, and a travel trajectory S5 that travels on a section of a single road that exits the intersection area Ac2.

Please refer to Fig. 2. The combination optimization unit 25 generates a new travel trajectory candidate by rearranging the connections between the nodes of multiple travel trajectories assigned to the same map link. Here, when rearranging the node connections, the connected nodes must be spatially smoothly connected. Therefore, the combination optimization unit 25 judges the connection possibility between the nodes as follows.
5, the position and orientation of node N1 are (x1, y1, θ1), and the position and orientation of node N2 are (x2, y2, θ2). The combination optimization unit 25 determines that node N1 can be connected to node N2 when all of the following connectable conditions (1), (2), and (3) are satisfied.

(x2-x1) ² + (y2-y1) ² <D ² ...(1)
atan((y2-y1)/(x2-x1))<A...(2)
|θ2−θ1|<B…(3)
The connectable conditions (1) and (2) mean that node N2 is located within the shaded area in FIG. 5 as viewed from node N1, and the connectable condition (3) means that the attitude of node N1 is similar to the attitude of node N2.

The combination optimization unit 25 generates new travel trajectory candidates from a plurality of travel trajectories assigned to the same map link based on these connection enabling conditions (1) to (3).
The procedure by which the combination optimization unit 25 generates candidate travel trajectories will be described with reference to FIG.
In step S1, the combination optimization unit 25 acquires travel routes assigned to the same map link. The total number of acquired travel routes is set to N.
In step S2, the combination optimization unit 25 randomly rearranges the order of the acquired travel trajectories and assigns them ranks 1 to N. The i-th travel trajectory is denoted as travel trajectory Ti (i: 1 to N).

In step S3, the combination optimization unit 25 initializes the value of a counter i to "1".
In step S4, the combination optimization unit 25 determines the connection possibility for all combinations of nodes of the travel trajectory Ti and the travel trajectory T(i+1). Here, for the combination of the travel trajectory Ti and the travel trajectory T(i+1), the combination optimization unit 25 separately determines the connection possibility when a node of the travel trajectory Ti is connected to a node of the travel trajectory T(i+1) and the connection possibility when a node of the travel trajectory T(i+1) is connected to a node of the travel trajectory Ti.

In step S5, the combination optimization unit 25 generates multiple driving trajectory candidates by randomly connecting connectable nodes (i.e., randomly rearranging the connections between the nodes of the driving trajectory Ti and the driving trajectory T(i+1)).
Please refer to Fig. 7. Travel trajectories T1 and T2 are travel trajectories acquired when traveling on a two-lane road. The white plots with reference signs N1, N2, ... N8 indicate nodes on the travel trajectory T1, and the white plots with reference signs Na, Nb, ... Ng indicate nodes on the travel trajectory T2. These travel trajectories T1 and T2 intersect at two points and can be connected at these intersections.

8 shows a schematic diagram of the connectivity between these nodes N1 to N8 and Na to Ng. In FIG. 8, two nodes connected by an arrow indicate that they can be connected in the direction indicated by the arrow. For example, node Nc of the travel locus T2 can be connected to node N4 of the travel locus T1, node N4 of the travel locus T1 can be connected to node Nd of the travel locus T2, and node Nf of the travel locus T2 can be connected to node N7 of the travel locus T1.
In this case, for example, the combination optimization unit 25 may generate a combination of travel trajectory candidates starting from node N1, proceeding from node N4 to node Nd, and then returning from node Nf to node N7 again. Alternatively, a travel trajectory candidate may be generated that proceeds from node Nf to node Ng without returning to node N7.
Alternatively, for example, a travel trajectory candidate may be generated that starts from node N2, proceeds from node Nc to node N4, and proceeds from node N4 to node N8. The combination optimization unit 25 may also randomly generate a plurality of travel trajectory candidates.

6. In step S6, the combination optimization unit 25 calculates an evaluation value for each of all the generated travel trajectory candidates and travel trajectories Ti and T(i+1) by adding up the reliability of all the nodes belonging to these trajectory candidates and travel trajectories. For example, the combination optimization unit 25 may calculate the sum of the reliability of all the nodes as the evaluation value, or may calculate the average value as the evaluation value.
In step S7, the combination optimization unit 25 selects one of these travel trajectory candidates and travel trajectories Ti and T(i+1) based on the evaluation value. For example, the travel trajectory candidate or travel trajectory having the largest evaluation value may be selected, or one of the travel trajectory candidates or travel trajectories having an evaluation value equal to or greater than a threshold value may be selected. The combination optimization unit 25 replaces (overwrites and updates) the travel trajectory Ti with the selected travel trajectory candidate or travel trajectory.

In step S8, the combination optimization unit 25 increments the value of the counter i by one.
In step S9, the combination optimization unit 25 determines whether the value of the counter i is equal to or greater than N. If the value of the counter i is equal to or greater than N (step S9: Y), the process proceeds to step S10. If the value of the counter i is less than N (step S9: N), the process returns to step S4.

In step S10, the combination optimization unit 25 determines whether the processes of steps S2 to S9 have been repeated a predetermined number of times. If the processes of steps S2 to S9 have not yet been repeated the predetermined number of times (step S10: N), the process returns to step S2. If the processes have been repeated the predetermined number of times (step S10: Y), the combination optimization unit 25 ends the process.
The travel trajectory Ti (i: 1 to N) updated through the above processing is obtained as a set of N travel trajectory candidates. The combination optimization unit 25 stores the set of travel trajectory candidates in the trajectory candidate storage unit 26.

These processes obtain a set of driving trajectory candidates (i.e., driving trajectory candidates that run on a driving line close to the center of the lane) that have a better evaluation value than the original driving trajectory obtained by manually driving the vehicle 1. However, these driving trajectory candidates include multiple driving trajectory candidates that run on the same lane, and are therefore redundant.
Here, a driving trajectory in which vehicles travel in the same lane includes, for example, on a multi-lane road as shown in Figure 7, two driving trajectories in which both vehicles travel only in the right lane, or in which both vehicles travel only in the left lane, as well as in which both vehicles change lanes from the right lane to the left lane, or in which both vehicles change lanes from the left lane to the right lane.

For this reason, the combination optimization unit 25 identifies multiple travel trajectory candidates that travel along the same lane.
For example, the combination optimization unit 25 may compare any two of the multiple travel trajectory candidates and determine whether or not these two travel trajectory candidates are spatially independent. If the two travel trajectory candidates are spatially independent, it may determine that they are not travel trajectory candidates traveling on the same lane, and if they are not independent, it may determine that they are travel trajectory candidates traveling on the same lane.

For example, the combination optimization unit 25 may determine whether or not two driving trajectory candidates are spatially independent based on the distance between the starting nodes and the distance between the end nodes of the two driving trajectory candidates.
9 , for example, the combination optimization unit 25 calculates a distance DS between node 1S, which is a start point of the traveling trajectory candidate T1, and node 2S, which is a start point of the traveling trajectory candidate T2, and a distance DE between node 1E, which is an end point of the traveling trajectory candidate T1, and node 2E, which is an end point of the traveling trajectory candidate T2.
When at least one of the distance DS and the distance DE is greater than a threshold, the combination optimization unit 25 determines that the travel trajectory candidates T1 and T2 are spatially independent, i.e., that the travel trajectory candidates T1 and T2 are not travel trajectory candidates that travel in the same lane.
On the other hand, if both the distances DS and DE are equal to or smaller than the thresholds, it is determined that the travel trajectory candidates T1 and T2 are not spatially independent, i.e., the travel trajectory candidates T1 and T2 are travelling in the same lane.

The combination optimization unit 25 selects one of the multiple travel trajectory candidates identified as a travel trajectory that travels along the same lane. For example, the combination optimization unit 25 may retain (extract) one of the multiple identified travel trajectory candidates and delete (discard) the rest from the trajectory candidate storage unit 26.
The combination optimization unit 25 may select the travel trajectory candidates based on, for example, the evaluation value. For example, the travel trajectory candidate having the largest evaluation value may be selected, or any of the travel trajectory candidates having an evaluation value equal to or greater than a predetermined value may be selected.

10. When at least one of the distance DS between the node 1S of the start point of the travel trajectory candidate T1 and the node 2S of the start point of the travel trajectory candidate T2 and the distance DE between the node 1E of the end point of the travel trajectory candidate T1 and the node 2E of the end point of the travel trajectory candidate T2 is greater than a threshold, it is determined that the travel trajectory candidates T1 and T2 are not travel trajectory candidates that travel in the same lane. Therefore, the combination optimization unit 25 does not perform a process of selecting one of the travel trajectory candidates T1 and T2. In other words, both the travel trajectory candidates T1 and T2 are retained.
On the other hand, if both the distance DS and the distance DE are equal to or smaller than the threshold, it is determined that the travel trajectory candidates T1 and T2 are travelling in the same lane. Therefore, the combination optimization unit 25 holds one of the travel trajectory candidates T1 and T2 with the larger evaluation value, and deletes the other with the smaller evaluation value from the trajectory candidate storage unit 26.

(Driving assistance mode)
Please refer to Fig. 2. In the driving assistance mode, the driving control device 7 assists the driving of the host vehicle 1 along the driving trajectory candidate generated in the trajectory candidate generation mode.
The road determination unit 27 identifies the map link of the road on which the vehicle 1 is traveling, based on the vehicle's own position acquired by the vehicle position acquisition device 4 .
The road shape estimation unit 28 acquires driving trajectory candidates assigned to the map links of the road being traveled from the trajectory candidate memory unit 26, selects one driving trajectory candidate having a node closest to the current position of the vehicle 1, and sets it as the target driving trajectory.

The vehicle control unit 29 controls the actuator 6 to assist the host vehicle 1 in traveling along the target traveling trajectory. For example, when autonomous traveling control is executed as the traveling assistance control, the vehicle control unit 29 controls the actuator 6 to automatically travel the host vehicle 1 along the target traveling trajectory.
Furthermore, for example, when driving assistance control is executed as driving assistance control, the actuator 6 is controlled so as to obtain a steering angle that causes the host vehicle 1 to travel along the target traveling trajectory. For example, a steering assist force may be applied so that the host vehicle 1 travels along the target traveling trajectory.

(Modification)
In the above description, the driving assistance device generates a target driving trajectory for driving the host vehicle 1 based on a driving trajectory traveled in the past by the host vehicle 1. Alternatively, a target driving trajectory for driving another vehicle other than the host vehicle 1 may be generated. In addition, a target driving trajectory for driving the host vehicle 1 may be generated based on a driving trajectory traveled in the past by another vehicle other than the host vehicle 1, and a target driving trajectory for driving the other vehicle or another vehicle may be generated.

(Effects of the embodiment)
(1) The driving control device 7 acquires driving trajectories when the vehicle drives a predetermined section of the same road multiple times, extracts multiple passing points, which are positions on the driving trajectory, for each acquired driving trajectory, assigns a reliability to each of the multiple passing points, and stores the passing points and the reliability in association with each other in the storage device.
By connecting the passing points stored in the storage device, a plurality of candidate driving trajectories for traveling through a specified section are generated, and one of the generated plurality of candidate driving trajectories is extracted based on a combined value of the reliability of the passing points belonging to the candidate driving trajectory, and the extracted candidate driving trajectory is generated as a target driving trajectory along which the vehicle will travel.
This makes it possible to set an optimized target driving trajectory based on the driving locus when the vehicle has traveled a predetermined section of the same road a number of times in the past.

(2) The driving control device 7 may calculate the distance of the passing point from the center of the lane of the road, and may assign a higher reliability to a case where the calculated distance is small than a case where the calculated distance is large. This allows a target driving trajectory optimized to drive near the center of the lane to be set.
(3) The driving control device 7 may estimate the center of the lane from the positions of lane boundary lines of the lane on the road. This allows the center of the lane to be estimated based on the recognition result of the lane boundary lines, and the reliability of the passing point to be calculated.

(4) The driving control device 7 may extract driving trajectory candidates whose sum is equal to or greater than a predetermined value. This makes it possible to set a target driving trajectory that is optimized to drive near the center of the lane.
(5) The driving control device 7 may generate multiple driving trajectory candidates by determining whether or not the passing points stored in the storage device are connectable, and connecting the passing points that are determined to be connectable.
For example, if the distance between two pass points is equal to or less than a distance threshold and the difference in the attitude angle or traveling direction of the vehicle at the two pass points is equal to or less than an angle threshold, it may be determined that the two pass points are connectable.
This makes it possible to generate a candidate travel trajectory in which passing points are smoothly connected.

(6) The driving control device 7 may generate a plurality of driving trajectory candidates by connecting passing points that belong to different driving trajectories among the acquired driving trajectories. This makes it possible to generate a driving trajectory candidate that is more optimized than a driving trajectory obtained by driving a predetermined section of the same road.
(7) The driving control device 7 may generate multiple driving trajectory candidates by repeating multiple times the process of extracting any two of the acquired driving trajectories and the process of generating a driving trajectory candidate by connecting a passing point belonging to one of the two driving trajectories with a passing point connecting to the other driving trajectory.
This allows a larger number of candidate driving trajectories to be generated, thereby making it possible to set a more optimized target driving trajectory.

(8) The driving control device 7 may compare any two of a plurality of driving trajectory candidates, and extract both of the two driving trajectory candidates if the two driving trajectory candidates are spatially independent, and extract one of the two driving trajectory candidates and discard the other if the extracted driving trajectory candidates are not spatially independent.
For example, it may be determined whether the two travel trajectory candidates are spatially independent based on the distance between the starting pass points and the distance between the end pass points of the two travel trajectory candidates.
This makes it possible to delete redundant travel trajectory candidates, thereby reducing the storage capacity for storing the travel trajectory candidates and the processing load for selecting the target travel trajectory from the travel trajectory candidates.
(9) The vehicle may be assisted in traveling along the target traveling trajectory generated by the above (1) to (11).
This makes it possible to assist the vehicle in driving based on the optimized target driving trajectory.

1... host vehicle, 2... vehicle speed sensor, 3... yaw rate sensor, 4... host vehicle position acquisition device, 5... lane recognition device, 6... actuator, 7... driving control device, 10... processor, 11... storage device, 20... driving trajectory memory unit, 21... trajectory optimization unit, 22... reliability calculation unit, 23... map information memory unit, 24... link assignment unit, 25... optimization unit, 26... trajectory candidate memory unit, 27... driving road determination unit, 28... road shape estimation unit, 29... vehicle control unit

Claims (11)

Acquire a travel trajectory of a vehicle traveling a predetermined section of the same road multiple times;
extracting a plurality of passing points, which are positions on the travel locus, for each of the acquired travel locus and assigning a reliability to each of the plurality of passing points;
storing the passing points and the reliability in a storage device in association with each other;
generating a plurality of candidate travel trajectories for travelling through the predetermined section by connecting the passing points stored in the storage device;
extracting one of the plurality of travel trajectory candidates based on a sum of the reliability of the passing points belonging to the travel trajectory candidate from the plurality of travel trajectory candidates that have been generated;
generating the extracted travel trajectory candidate as a target travel trajectory along which the vehicle will travel;
calculating a distance of the passing point from a center of a lane of the road, and assigning a higher reliability when the calculated distance is small than when the calculated distance is large ;
A traveling trajectory generating method comprising: 2. The method according to claim 1 , further comprising estimating a center of the lane from positions of lane boundary lines of the lane on the road. 3. The traveling trajectory generating method according to claim 1, further comprising extracting the traveling trajectory candidates whose sum is equal to or greater than a predetermined value. The travel trajectory generation method according to any one of claims 1 to 3, characterized in that the plurality of travel trajectory candidates are generated by determining whether the pass points stored in the storage device are connectable or not , and connecting the pass points determined to be connectable. 5. The driving trajectory generation method according to claim 4, characterized in that it is determined that the two passing points are connectable when a distance between the two passing points is equal to or less than a distance threshold and a difference in an attitude angle or a traveling direction of the vehicle at the two passing points is equal to or less than an angle threshold. The travel trajectory generating method according to any one of claims 1 to 5 , characterized in that the plurality of travel trajectory candidates are generated by connecting the passing points belonging to different travel trajectories among the acquired travel trajectories. A process of extracting any two of the acquired travel trajectories;
generating the travel trajectory candidate by connecting a passing point belonging to one of the two travel trajectories with a passing point connected to the other travel trajectory;
The traveling trajectory generating method according to claim 6 , wherein the plurality of traveling trajectory candidates are generated by repeating the above a plurality of times. The method for generating a travel trajectory according to any one of claims 1 to 7, characterized in that any two of the plurality of travel trajectory candidates are compared, and when at least one of a first distance between the pass points of start points of the two travel trajectory candidates or a second distance between the pass points of end points of the two travel trajectory candidates is greater than a threshold, both of the two travel trajectory candidates are extracted, and when both of the first distance and the second distance are equal to or less than the threshold, one of the two travel trajectory candidates is extracted and the other is discarded. A driving support method, comprising: supporting the vehicle to drive along the target driving trajectory generated by the driving trajectory generation method according to any one of claims 1 to 8 . A positioning device for measuring the current position of a vehicle;
A vehicle sensor for detecting a running state of the vehicle;
A storage device;
a controller that acquires a travel trajectory when the vehicle has traveled a predetermined section of the same road a plurality of times based on a positioning result of the positioning device and a detection result of the vehicle sensor, extracts a plurality of passing points which are positions on the travel trajectory for each of the acquired travel trajectories, assigns a reliability to each of the plurality of passing points, associates the passing points with the reliability and stores them in the storage device, generates a plurality of travel trajectory candidates for travelling the predetermined section by connecting the passing points stored in the storage device, extracts one of the plurality of generated travel trajectory candidates based on a sum of the reliability of the passing points belonging to the travel trajectory candidate, generates the extracted travel trajectory candidate as a target travel trajectory on which the vehicle will travel, calculates a distance of the passing point from a center of a lane of the road, and assigns a higher reliability when the calculated distance is small than when the calculated distance is large ;
A running trajectory generating device comprising: A driving assistance device that assists the vehicle in driving along the target driving trajectory generated by the driving trajectory generation device according to claim 10 .

Images