How to Plan Lane Change Maneuvers for Autonomous Vehicles

Abstract

The invention relates to a method for planning a lane change maneuver from one driving lane (F1, F2) to an outer driving lane (F3) for an automated vehicle (1) in preparation for exiting a multi-lane road section (F), said vehicle (1) being connected or connectable in data-technical terms to a central computing unit (3). According to the invention, the central computing unit (3) receives traffic data representative of the local traffic volume on the multi-lane road section (F), receives lane change data from another automated vehicle of a fleet of vehicles relating to a planned and successfully executed automated lane change maneuver, and the central computing unit (3) determines, based on the received traffic data and the received lane change data, a probability that an automated lane change maneuver of the vehicle (1) can be executed at a specific position in the current traffic volume.

Description

[Technical field]

[0001] The present invention relates to a method for planning a lane change maneuver from one driving lane to an outer driving lane in preparation for exiting a multi-lane road section for an autonomous vehicle, in which the vehicle is connected or can be connected in data-technical terms to a central computing unit. [Background technology]

[0002] German Patent No. DE 102019005446 discloses a method for the automated merging of a vehicle from a merging lane into a driving lane. In this method, a primary merging position is determined in advance, at which the vehicle is to change from the merging lane into the driving lane. If a lane change at the primary merging position or in the vicinity of the primary merging position does not disturb road users present in the driving lane, the vehicle is controlled so that a lane change is performed at the primary merging position. If a lane change is not performed at the primary merging position, a lane change is performed at a secondary merging position, which is determined depending on the detected free space between road users present in the driving lane.

[0003] Furthermore, German Patent Application No. 10 2017221104 A1 describes a method for adjusting at least one distance between at least two autonomous or semi-autonomous vehicles in a vehicle group, in which at least one entrance and / or exit of a road on which the at least two vehicles are traveling in an area in front of the vehicle group is recorded, a situation-dependent optimal distance between the at least two vehicles when passing the entrance and / or exit is determined, the distance between the at least two vehicles is adapted to the situation-dependent optimal distance between the at least two vehicles before reaching the entrance and / or exit, the entrance and / or exit is passed with the determined situation-dependent optimal distance between the at least two vehicles, the situation-dependent optimal distance between the at least two vehicles is determined again after passing the entrance and / or exit, and the distance between the at least two vehicles is adapted to the newly determined situation-dependent optimal distance between the at least two vehicles after passing the entrance and / or exit. Summary of the Invention [Problem to be solved by the invention]

[0004] It is an object of the present invention to provide a method for planning lane change maneuvers for an autonomous vehicle. [Means for solving the problem]

[0005] This object is achieved according to the invention by a method having the features of claim 1.

[0006] Advantageous embodiments of the invention are the subject matter of the dependent claims.

[0007] According to the invention, a method for planning a lane change maneuver from one driving lane to an outer driving lane for an automatically driven vehicle in preparation for leaving a multi-lane road section (approaching an exit from a multi-lane road section) is provided, the method comprising the steps of: the vehicle being connected or connectable in data technology to a central computing unit, by which traffic data representative of the local traffic volume on the multi-lane road section is received; additionally, the central computing unit receives lane change data from another automatically driven vehicle of the vehicle fleet, relating to a planned and successfully executed automated lane change maneuver; and, based on the received traffic data and the received lane change data, the central computing unit determines a probability that an automated lane change maneuver of the vehicle can be executed at a specific position in the current traffic volume.

[0008] In this context, the outer driving lane is understood to mean the right driving lane in case of right-hand traffic and the left driving lane in case of left-hand traffic, in particular on highways.

[0009] That is, by applying this method, a location is determined for an autonomous vehicle to initiate a lane change maneuver to change lanes into an outer driving lane, thereby enabling exit from a highway while complying with legal requirements and technical constraints.

[0010] By using this method, the traffic flow can be improved because the autonomous vehicles perform the lane change maneuver in a timely manner to exit the respective highway, so that each autonomous vehicle does not have to recognize a relatively short time interval (time gap) or the vehicle is not required to be operated manually by the occupant, i.e. the driver, in order to reach the highway exit in a timely manner. When this method is applied, the traffic safety and driving comfort can be increased.

[0011] Compared to solution approaches based on vehicle-to-vehicle communication, the method according to the invention is substantially independent of the behavior of other road users, thus reducing both the technical complexity and the dependency on the cooperation of other vehicle manufacturers and / or road users.

[0012] In particular, by applying the method, it becomes easier for an autonomous vehicle to exit a highway at an exit, for example by merging early into an outer lane. Such a relatively early merging position into the outer lane can be determined, for example, from a fleet of autonomous vehicles that were able to exit the highway via the exit without driver intervention. Furthermore, the location where a human driver would normally change lanes into the outer lane to exit the highway at the exit can be determined from the fleet. Due to technical constraints of autonomous vehicles, the position for initiating the lane change maneuver of the autonomous vehicle can also be determined at a parameterizable longer distance, for example 300 m before the exit.

[0013] In other words, the method is based on the spatio-temporal investigation of the lane change maneuvers at the exit of the highway of another networked automated driving vehicle with the services of the central computing unit, which significantly reduces the risk of overshooting the exit and significantly increases the safety and driving comfort during the lane change maneuvers. Based on the lane change data and traffic data, the time to perform the lane change maneuver can be dynamically determined.

[0014] In one embodiment of the method, the lane change data transmitted from another autonomous vehicle of the fleet to the central computing unit includes lane change position information and lane change success information transmitted from another autonomous vehicle of the fleet to the central computing unit. Thus, the lane change data of the other vehicle contains sufficient information for the autonomous vehicle to determine a suitable position for starting a lane change maneuver, so that the vehicle can take the next exit with little risk. In particular, the position is determined based on the current lane change data in such a way that the lane change maneuver is not started suddenly, but gradually and in a manner adapted to the traffic situation, so that the driving comfort for the vehicle occupants before and during the lane change maneuver is not impaired.

[0015] In one embodiment of the method, the lane change position information is used to identify a location where a lane change maneuver was initiated for each other autonomous vehicle in the fleet, i.e., the lane change data includes information regarding the exact location where each other autonomous vehicle initiated a lane change maneuver to change from a current driving lane to an outer driving lane.

[0016] In another embodiment, the lane change success information is used to indicate whether the execution of the lane change maneuver of each other autonomous vehicle of the fleet was successful or had to be aborted, for example because a takeover request was output to the driver regarding the vehicle's driving task. The position for starting or commencing the lane change maneuver, in particular the position detected by satellite aiding, can be determined based on the calculation of the spatial distribution of successfully executed lane change maneuvers. In particular, before exiting the highway, a position is determined where the automated lane change maneuver can be successful (successfully executed) with a parameterizable probability.

[0017] In one further embodiment of the method, the start of a lane change maneuver is identified by actuation of a turn signal of the autonomous vehicle, where actuation of the turn signal serves as the beginning of the lane change maneuver and triggers transmission of lane change data to the central computing unit, and the start of a lane change maneuver of each of the autonomous vehicles is also identified based on actuation of the turn signal.

[0018] For this purpose, in one possible embodiment, the current position of the vehicle and each further vehicle at the time of activation of the turn signal is detected and transmitted to a central computing unit, said current position and time being taken as the spatiotemporal position of the start of the calculated lane change maneuver of the autonomous vehicle.

[0019] In one possible embodiment of the method, the traffic data is transmitted by the traffic and / or navigation service, in particular continuously, at least at regular time intervals, to the central computing unit, which is thus constantly provided with relatively current traffic data, which are used to determine (determine, calculate) a position at which a lane change maneuver can be performed in order to take the next exit.

[0020] Furthermore, in one development of the method, when determining whether the vehicle is capable of performing an automated lane change maneuver at a particular location and in a given current traffic volume, the method further comprises: - Weather conditions around the vehicle, - Visibility conditions around the vehicle, - Blockades around the vehicle, and / or - Dynamic traffic information Therefore, the specific position is adapted according to the surrounding conditions of the vehicle so that a safe and comfortable lane change maneuver can be performed.

[0021] Further, in one embodiment, based on the vehicle fleet data, capacity requirements are determined for the expected (probable) exit process specific to each exit, and the position for initiating the lane change maneuver is adapted (adjusted) according to the determined (determined) capacity requirements.

[0022] In the following, an embodiment of the invention will be explained in detail with reference to the drawings. [Brief description of the drawings]

[0023] [Figure 1] 1 shows a schematic diagram of a roadway section of a highway in the area of ​​an exit. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0024] The only figure shows a plan view of the roadway section F of the motorway in the area of ​​exit A.

[0025] The roadway section F includes an inside driving lane F1, a center driving lane F2, and an outside driving lane F3, which is the right-hand driving lane in cases of right-hand traffic and the left-hand driving lane in cases of left-hand traffic.

[0026] A vehicle 1 traveling in an automatic driving mode, in particular an autonomous driving mode, travels in the central driving lane F2, while a train of multiple large freight vehicles 2 travels in the outer driving lane F3.

[0027] When such large freight vehicles 2 travel relatively close to one another and the traffic volume is relatively high, e.g. during the evening rush hour and / or traffic in metropolitan areas, the possibilities for performing a lane change maneuver of the vehicle 1 from the central travel lane F2 to the outer travel lane F3 are often limited, so that the vehicle 1 may not be able to exit the highway from this exit A. In particular, in situations where the gradient of the exit A is relatively steep, a lane change maneuver is hardly possible or not possible at all.

[0028] A vehicle 1 traveling in an autonomous driving mode needs a sufficiently long time interval to perform a lane change maneuver relatively safely and comfortably. Since multiple large freight vehicles 2 are traveling in a platoon close to each other, there is no sufficiently long time interval.

[0029] In order to enable an autonomous vehicle 1 to perform lane change maneuvers on a highway in a relatively safe and comfortable manner, in particular for exiting the highway, a solution is described below.

[0030] The autonomous vehicle 1 as well as further vehicles not shown in detail belong, for example, to a common vehicle fleet of a vehicle manufacturer and are data-technically (from a data technology point of view) connected to a central computing unit 3.

[0031] At regular time intervals, the vehicle 1 as well as the other vehicles transmit their respective current positions, including the driving lane F1-F3 in which they are currently traveling, to the central computing unit 3.

[0032] Furthermore, the central computing unit 3 is connected in data technology to a navigation service 4, which can alternatively or additionally also be connected to a traffic service. The central computing unit 3 receives traffic data which represent the respective local traffic volume, for example for a roadway section F.

[0033] Based on the spatio-temporal investigation of lane change maneuvers at an exit A of a highway of another autonomous vehicle connected to the central computing unit 3 by a network, the method can significantly reduce the risk of missing the exit and can increase the safety and driving comfort when performing lane change maneuvers. In particular, the method can dynamically determine the spatio-temporal position for lane change maneuver of the autonomous vehicle 1 based on available lane change data and traffic data. The lane change data of the other vehicle includes lane change position information and lane change success information.

[0034] For each automatically driving vehicle 1, for example, an exit from the highway via exit A can be realized by the vehicle 1 merging into the outer driving lane F3 at a comparatively early time. For example, such an early position for initiating a lane change maneuver from the central driving lane F2 to the outer driving lane F3 can be detected (recorded) based on lane change data transmitted by another automatically driving vehicle to the central computing unit 3. In particular, for this purpose, lane change maneuvers are taken into account in which the other vehicle successfully exited the highway via exit A without driver intervention.

[0035] Furthermore, based on the data transmitted from the fleet, the position at which the occupant, i.e., the driver, of vehicle 1 would normally change lanes into the outer driving lane F3 to use exit A and thus initiate an automated lane change maneuver can be determined.

[0036] Due to technical constraints of the autonomous vehicle 1, the position for initiating a lane change maneuver can also be set at a parameterizable larger distance, e.g. 300 m to exit A.

[0037] To carry out the method, the central computing unit 3 is provided with the destination of the autonomous vehicle 1, so that it knows where the vehicle 1 must leave the highway. The central computing unit 3 is further provided with traffic data, as described above, which represent the traffic volume in the traffic network and thus also the traffic volume on the roadway section F.

[0038] Furthermore, the central computing unit 3 receives lane change data from the other autonomous vehicles of the fleet, relating to planned and successfully executed automated lane change maneuvers. The start of the respective lane change maneuver is identified in particular by the activation (setting) of the direction indicator of the respective autonomous vehicle. In particular, the position at which the lane change maneuver of the respective autonomous vehicle is started is transmitted to the central computing unit 3 by means of the lane change data as lane change position information. The activation of the direction indicator indicates to road users around the respective autonomous vehicle that a lane change maneuver is about to be executed.

[0039] The successful execution of a lane change maneuver by another vehicle in the area of ​​exit A, i.e. at exit A, is transmitted by this other vehicle to the central computing unit 3 as lane change success information, respectively. The lane change success information is a component of the lane change data.

[0040] Based on the current traffic data and the lane change data of other vehicles, a probability is then determined (calculated) that the vehicle can perform an automated lane change maneuver at a specific location, for example at a predefined distance to the exit A. In other words, based on the spatio-temporal distribution of successful execution of lane change maneuvers, a position for starting the lane change maneuver of the vehicle 1 is determined, calculated with satellite assistance, at which a lane change maneuver is successfully executed with a parameterizable probability before the exit A of the highway.

[0041] The location where an unsuccessful lane change maneuver was initiated can also be used to determine the location for a future lane change maneuver, for example by initiating the lane change maneuver 50 m earlier. This approach is particularly useful when there is no successful lane change information available yet.

[0042] In an alternative embodiment, data from a fleet of vehicles at an exit A of the highway can determine the most likely (probable) merging position into the outer driving lane F3, which can be extended for the autonomous vehicle 1 by a parameterizable distance to the exit, for example 300 m.

[0043] In one possible embodiment of the method, the parameterization of the probability and the merging position can be performed depending on the current traffic volume. That is, if the traffic volume is relatively very high, it is reasonable to start the lane change maneuver earlier in time. This can be done, for example, depending on the traffic phase, in particular according to the three-phase traffic theory. For example, in free traffic, for example 200 m before exit A, in synchronized traffic due to increased traffic density, for example 600 m before exit A, and in case of widespread congestion as a traffic phase, since in this case longer time gaps sometimes occur randomly, the lane change maneuver is started, for example 400 m before exit A.

[0044] Based on the vehicle fleet data and also based on, for example, lane change data, a capacity requirement (capacity demand, capacity condition, capacity requirement) for a prospective (anticipated) exit maneuver is determined specifically for each exit A. In case of spatiotemporal increased capacity requirements, for example in evening rush hour traffic and / or traffic at the start of a holiday and at the end of a holiday, the location for initiating the lane change maneuver can be shifted upstream by a parameterizable distance. Thus, a controlled traffic flow can be achieved. Furthermore, in order to exit the highway in a safe and comfortable manner, other parameters, for example current weather conditions, visibility conditions, adjacent lane closures, and / or dynamic traffic information, can be taken into account when determining the location for initiating the lane change maneuver.

[0045] The application of the method is particularly suitable for vehicle 1 of a fleet of vehicles, where information can be provided, if necessary and for a fee, to other fleet vehicles to perform lane change maneuvers on the highway in order to exit the highway. [Prior art documents] [Patent documents]

[0046] [Patent Document 1] German Patent No. 102019005446 [Patent Document 2] DE 102017221104

Claims

1. 1. A method for planning a lane change maneuver from one driving lane (F1, F2) to an outer driving lane (F3) for an exit from a multi-lane road section (F) for an autonomous vehicle (1), said vehicle (1) being connected or connectable in data-technical terms to a central computing unit (3), - said central computing unit (3) receives traffic data representative of local traffic volumes on said multi-lane roadway section (F); - receiving, by said central computing unit (3), lane change data relating to a planned and successfully executed automated lane change maneuver from another autonomous vehicle of the fleet; - the central computing unit (3) determines the probability that an automated lane change maneuver of the vehicle (1) can be performed at a particular position in the current traffic flow based on the received traffic data and the received lane change data; A method comprising:

2. The lane change data transmitted from the other autonomous vehicles of the fleet to the central computing unit (3) includes lane change position information and lane change success information.

2. The method according to claim 1 .

3. The lane change position information is used to identify a position where the lane change maneuver of the other autonomous vehicle in the fleet was initiated.

3. The method according to claim 2 .

4. The lane change success information is used to indicate whether the lane change maneuver of each of the other autonomous vehicles in the fleet was successful or had to be aborted.

4. The method according to claim 2 or 3.

5. The initiation of a lane change maneuver is identified based on the activation of a turn signal of the autonomous vehicle (1). The method according to any one of claims 1 to 4, characterized in that

6. The current position of the autonomous vehicle (1) at the time of activation of the turn signal is detected and transmitted to the central computing unit (3).

6. The method according to claim 5 .

7. The traffic data is transmitted to the central computing unit (3) by a traffic service and / or a navigation service (4). The method according to any one of claims 1 to 6, characterized in that

8. when determining whether the automated lane change maneuver of the vehicle (1) can be performed at a particular location in a current traffic flow, - the weather conditions around the vehicle, - visibility conditions around the vehicle, - a seal around the vehicle, and / or - Dynamic traffic information is taken into consideration The method according to any one of claims 1 to 7, characterized in that

9. Based on the fleet data, capacity requirements for the expected exit process specific to each exit (A) are determined. The method according to any one of claims 1 to 8, characterized in that

10. A position for the lane change maneuver is adapted depending on the determined capacity requirement.

10. The method according to claim 9.

Images