Advanced highway assist scenario

Abstract

A multi-lane scenario driving support is provided for an ego vehicle in a traffic situation. Traffic surroundings are measured by an environment sensor system. The traffic surroundings include data about traffic and free space within an ego lane of the ego vehicle and an adjacent lane, and data about front proximity area and rear proximity area of the ego vehicle. A decision device evaluates the measured traffic surroundings and decides a driving operation to be executed by the ego vehicle based on at least one strategy. In the decision device, a cost function is used for choosing one of at least six strategies. The cost function is based on at least a core priority to avoid collision of the ego vehicle and not cause collision of the ego vehicle with a third party vehicle.

Claims

1. A method for providing a multi-lane scenario driving support for an ego vehicle in a traffic situation in which an ego lane is partly or fully obstructed by other traffic participants, the method comprising: measuring traffic surroundings by an environment sensor system, wherein the traffic surroundings include data about traffic and free space within the ego lane of the ego vehicle and at least one adjacent lane, and data about front proximity area and rear proximity area of the ego vehicle; evaluating, by a decision device the measured traffic surroundings together with the speed of the ego vehicle and deciding a driving operation to be executed by the ego vehicle based on at least one strategy, wherein in the decision device a cost function is used for choosing one of at least six strategies, the cost function being based on at least a core priority, whereby the core priority is to avoid collision of the ego vehicle and not cause collision of the ego vehicle with a third party vehicle, whereby the decision device by the cost function chooses one selected from the group consisting of at least the following six strategies: braking in the ego lane, to combine braking and steering within the ego lane of the ego vehicle, steering within the ego lane of the ego vehicle to avoid an obstacle, to full-brake in the ego lane of the ego vehicle, to combine braking and steering towards or when entering temporarily the adjacent lane and steering towards or when entering temporarily the adjacent lane; wherein the cost function is used in mathematical optimization and decision theory for mapping an event, finding a real number representing a cost associated with the event, and choosing the one of the at least six strategies with a lowest cost; and wherein the decision device by the cost function chooses the strategies in the following order: braking in the ego lane, to combine braking and steering within the ego lane of the ego vehicle, steering within the ego lane of the ego vehicle to avoid the obstacle, to full-brake in the ego lane of the ego vehicle, to combine braking and steering towards or when entering temporarily the adjacent lane, and steering towards or when entering temporarily the adjacent lane.

2. The method according to claim 1, wherein the cost function for choosing one of the at least six strategies is based on a core priority parameter, and assuming that core priority is guaranteed, at least one of the following additional parameters: minimum lateral acceleration of the ego vehicle, minimum longitudinal acceleration of the ego vehicle, and avoid leaving the ego lane.

3. The method according to claim 1, wherein a first strategy to gain the core priority is only braking in the ego lane.

4. The method according to claim 3, wherein the braking in the ego lane happens with a deceleration up to and including ?5 m/s.sup.2.

5. The method according to claim 3, wherein a deceleration limit varies based on selected driving mode.

6. The method according to claim 3, wherein a second strategy to gain the core priority is to combine braking and steering within the ego lane of the ego vehicle, whereby this second strategy is a strategy subordinate to the first strategy of only braking in the ego lane.

7. The method according to claim 6, wherein a third strategy to gain the core priority is only steering within the ego lane of the ego vehicle to avoid the obstacle, whereby the third strategy is a strategy subordinate to the second strategy to combine braking and steering within the ego lane of the ego vehicle.

8. The method according to claim 7, wherein the prioritization between the second strategy to combine braking and steering within the ego lane of the ego vehicle and the third strategy of only steering within the ego lane of the ego vehicle to avoid the obstacle is based on the relative speed between the ego vehicle and the obstacle.

9. The method according to claim 7, wherein a fourth strategy to gain the core priority is to only full-brake in the ego lane of the ego vehicle, wherein the fourth strategy is a strategy subordinate to the third strategy of only steering within the ego lane of the ego vehicle to avoid the obstacle.

10. The method according to claim 9, wherein a fifth strategy to gain the core priority is to combine braking and steering towards or when entering temporarily the adjacent lane, wherein the fifth strategy is a strategy subordinate to the fourth strategy to only full-brake in the ego lane of the ego vehicle.

11. The method according to claim 10, wherein a sixth strategy to gain the core priority is only steering towards or when entering temporarily the adjacent lane, wherein the sixth strategy is a strategy subordinate to the fifth strategy to combine braking and steering towards/entering temporarily the adjacent lane.

12. The method according to claim 1, wherein the decision device takes data about map information and a planned driving activity into account.

13. A system for providing the multi-lane scenario driving support for the ego vehicle in the traffic situation, using the method according to claim 1, the system comprising: the environment sensor system and the decision device.

14. The system according to the claim 13, wherein the environment sensor system includes a front camera, a radar cocoon and/or a laser scanner cocoon.

Description

(1) In the drawings:

(2) FIG. 1 shows a schematic top view of a critical traffic situation on a multi-lane road to be solved by a preferred embodiment of the invention,

(3) FIG. 2 shows a schematic top view of a critical traffic situation on a multi-lane road to be solved by the preferred embodiment of the invention similar to FIG. 1 and

(4) FIG. 3 shows a schematic top view of a critical traffic situation on a multi-lane road to be solved by the preferred embodiment of the invention similar to FIGS. 1 and 2.

(5) FIG. 1 shows a traffic situation in which the inventive method is applied. In particular, the present invention provides a method for providing a multi-lane scenario driving support for an ego vehicle 10 in a traffic situation. As shown more precisely, the ego vehicle is driving on its lane, which is also called ego lane 16. The wording ego vehicle means the driver's own vehicle, especially a car or truck. The wording ego lane means the lane on which the ego vehicle 10 drives. As can be seen in FIG. 1, the multi-lane road includes the case when a two-lane road becomes a three-lane road.

(6) The traffic surroundings are measured by an environment sensor system 14, which is shown as a part of the ego vehicle 10. The traffic surroundings include data about traffic and free space within the ego lane 16 of the ego vehicle 10 and at least an adjacent lane 12a, 12b. Present at FIG. 1, there are two adjacent lanes to each side of the ego vehicle 10. Data is measured about front proximity area 18 and rear proximity area 20 of the ego vehicle 10, whereby a decision device 22 evaluates the measured traffic surroundings together with the speed of the ego vehicle 10 and decides a driving operation to be executed by the ego vehicle 10 based on at least one (driving) strategy.

(7) The decision device 22 takes data about map information and a planned driving activity or driving operation into account. The map is used to track the other traffic participant(s), e.g. third party vehicle(s), such as third party vehicle 24. The map is also used to define the free space, meaning the area where there is most probably no other traffic participant or object.

(8) The decision device 22 decides the driving operation when the third party vehicle 24 is detected in the ego lane 16 at a specific distance. The decision device 22 decides the driving operation continuously (e.g. in each cycle) while the ego vehicle 10 is driving. This includes in particular the case when it is predicted (by using or evaluating the measured traffic surroundings) that an adjacent lane 12b is formed out of the ego lane 16, as shown in FIG. 1.

(9) In the decision device 22 a cost function is used/implemented for choosing one of at least six strategies. The cost function is based on at least a core priority. It is the core priority to avoid collision and not cause collision of the ego vehicle 10 with a third party vehicle 24. In FIG. 1 the third party vehicle can be the vehicle 24 in the ego lane and/or a vehicle in an adjacent lane 12a, 12b. By means of the cost function the decision device 22 choses one of at least six strategies to gain the core priority. These six strategies are (only) braking in the ego lane 16, to combine braking and steering within the ego lane 16 of the ego vehicle 10, (only) steering within the ego lane 16 of the ego vehicle 10 to avoid an obstacle, to (only) full-brake in the ego lane 16 of the ego vehicle 10, to combine braking and steering towards or when entering temporarily an adjacent lane 12a, 12b and steering towards or when entering temporarily an adjacent lane 12a, 12b.

(10) The cost function can be based not only on the core priority, but also on at least one additional parameter, assuming that core priority is guaranteed. The additional parameters can for example be: minimum lateral acceleration of the ego vehicle 10, minimum longitudinal acceleration of the ego vehicle 10, and avoid leaving the ego lane 16 (meaning no ego lane departure).

(11) FIG. 1 also shows that a new lane of the ego lane 16 is opened in the driving direction of the ego vehicle 10. It corresponds to the teaching of the invention that the ego vehicle 10 or its decision device 22 selects from one of the programmed strategies to fulfill the core priority, in order to grant the driver of the ego vehicle 10 a safe and at the same time comfortable ride.

(12) The basic idea of the invention is to gain an opportunity to monitor sufficiently an ego lane, further road traffic and the congestion situation. It turns out that these are the main strategies to avoid collision and not cause collision. It also allows in particular to handle the situation of ego lane 16 being partly occupied by a third party vehicle 24 or any other object. This means, the invention concerns a method for an at least semi-autonomous ego vehicle 10 on a multi-lane road. It is monitored whether the adjacent lanes 12a, 12b are free. If there is an obstacle in the lane 16 of the ego vehicle 10, as for example third party vehicle 24, the decision device 22 decides, depending on the occupancy of the adjacent lanes 12a, 12b, what driving operation the ego vehicle should execute, e.g. whether the ego vehicle 10 brakes, dodges or changes lanes.

(13) A main benefit of the above-mentioned method is the created possibility for intuitive driving in a complex traffic scenario. Further, there is a safe system reaction avoiding collision in a complex traffic scenario. Preferably, without such method the driver shall respond and take back the control quickly. As a consequence, the confidence of the end user is increased into the system of the ego vehicle. Thus, a basis is created towards higher level of automation where less driver attention and less driver actions are needed.

(14) This method shall be applied preferably to a level 2 or level 3 system which combines longitudinal and lateral control.

(15) While the ego vehicle 10 drives on the ego lane 16, the environment sensor system 14 (including sensors) monitors the front proximity area 18 and the rear proximity area 20. According to the situation shown in FIG. 1, it can be seen that a new adjacent lane 12b forms in front of the ego vehicle 10. A third party vehicle 24 is partly on the ego lane 16 and partly on the adjacent lane 12b in order to integrate into the adjacent lane 12b. The basic assumption is that the ego vehicle 10 approaches the aforementioned third party vehicle 24 (e.g. third party vehicle 24 is detected in the ego lane 16 at a specific distance) and that the system for providing a multi-lane scenario driving support for an ego vehicle 10 in a traffic situation is intended to apply the logic according to doctrine of the invention. For this the ego vehicle 10 with its environment sensor system 14 monitors the traffic situation surrounding the ego vehicle 10. In particular, redundant and cumulative sensors can be used to exclude possible measurement errors and thus increase driver safety and comfort. The decision device 22 processes this data and takes map information into account. This means that even if it is not yet possible to detect a new adjacent lane 12b using the sensors, it can be predicted that it will be formed. If available, it may also be possible to use data from an ongoing vehicle navigation system to further improve the prediction quality and thus solve the object of the invention as reliably as possible.

(16) In any case, it is the core priority of the system and the applied method, to avoid a collision and not cause a collision. The decision device 22 by means of the cost function chooses one of at least six strategies to gain the core priority. The core priority is achieved particularly preferably by a following order of the individual strategies, whereby this represents only a preferred arrangement of the invention. In principle, it is also possible to sort individual strategy points differently, depending on the cost function.

(17) According to a preferred form of the invention, the ego vehicle 10 will first try to (only) brake in the ego lane 16. This first strategy or step should take place as early as possible and preferably with a maximum deceleration of ?5 m/s.sup.2 in order to give the driver an optimal feeling of safety and driving comfort. Ideally, the third party vehicle 24 is fully integrated into the new adjacent lane 12b when the ego vehicle 10 passes by.

(18) If the decision device 22 concludes that the above-mentioned third party vehicle 24 is so far in the ego lane 16 that braking alone is not sufficient to avoid a collision, braking and steering within the ego lane 16 is preferably combined as the next (second) strategy or step, so that the ego vehicle 10 circumvents the third party vehicle 24 within the ego lane 16 sufficiently carefully.

(19) Some situations, though, may arise in which braking together with steering is not sufficient for reasons of danger, e.g. if the ego lane 16 should quickly move away from a danger point, or if braking would limit the driver's driving comfort, e.g. due to a gradient. In this case, (only) steering within the ego lane 16 of the ego vehicle 10 is intended to avoid an obstacle, in particular as a third strategy or step.

(20) However, it becomes risky if the third party vehicle 24 is further away than in FIG. 1 in ego lane 16 or if the vehicle shown in FIG. 1 in adjacent lane 12a suddenly moves in front of ego vehicle 10 in the direction of adjacent lane 12b and blocks ego lane 16. Depending on the surrounding traffic situation and the distance to the third party vehicle, the above mentioned strategies may not be sufficient to reach the core priority. In this example, (only) full-braking in the ego lane 16 of the ego vehicle 10 is preferred, in particular as a fourth strategy or step.

(21) If the braking distance for the ego vehicle 10 is not sufficient to prevent a collision with respect to the above exemplary scenario, a next (fifth) strategy or step could be to combine braking and steering towards or when entering temporarily an adjacent lane 12a, 12b. This strategy is of secondary importance to the one mentioned above because an evasive maneuver into an adjacent lane 12a, 12b is associated with a risk for other road users.

(22) Depending on the hazard situation, the combination of braking and steering towards or when entering temporarily an adjacent lane 12a, 12b may not be sufficient in the above-mentioned exemplary scenario, so that braking is then omitted and only one steering towards or when entering temporarily an adjacent lane 12a, 12b takes place (sixth strategy or step). However, this is the last preferred and yet effective strategy, because an unchecked lane change could endanger or at least irritate the surrounding road users. For this, to perform any lateral action, which requires entering the adjacent lane, has as requirement that lane change conditions are checked. Hence, rear and side detections may be mandatory requirements of this strategy.

(23) In principle, according to the doctrine of the invention, the preferred strategy for the above-mentioned strategies is always to follow a strategy only if the strategies for achieving the core priority mentioned in this strategy are not applicable as determined by the decision device 22.

(24) FIG. 2 shows a situation which is critical and challenging for ADAS according to a state of the art ego vehicle 10 on its ego lane 16, when a newly created, adjacent lane 12a is full of standstill of third party vehicles. In this context, a driving situation threatens to become confusing. The ego vehicle 10 is at risk of performing abrupt or unsafe driving maneuvers due to overlapping or contradictory specifications, when vehicles drive into the new lane.

(25) FIG. 3 shows another situation which is critical and challenging for ADAS according to a state of the art ego vehicle 10 on its ego lane 16, when standstill vehicles cross two lanes at the same time, namely a newly created, adjacent lane 12a and the ego lane 16.

(26) The critical traffic situations shown in the FIGS. 2 and 3 are not exhaustive, but merely exemplary. In reality, there are other critical traffic situations that are not described here.

(27) A most preferred, first strategy to gain the core priority is (only) braking in the ego lane 16. This feature causes the least discomfort to the driver and has the least risk to the vehicles surrounding the ego vehicle.

(28) In case of braking in the ego lane, it is preferred that it happens with a deceleration up to and including ?5 m/s.sup.2. This maximum value of deceleration makes the vehicle behavior appear as inconspicuous as possible and resembles the braking behavior of a driver without automated driving.

(29) If the strategy of (only) braking in the ego lane is chosen by the decision device, the deceleration limit can vary based on selected driving mode. In particular, the driving modes can be eco, comfort and sports, whereby the fastest deceleration might be achieved by the sports mode. The slowest deceleration might be achieved by the eco or the comfort mode. This depends on the environmental conditions, which can have an influence on the most economical way of driving.

(30) A further, second strategy to gain the core priority is to combine braking and steering within the ego lane 16 of the ego vehicle 10, whereby this second strategy or step is preferably a strategy subordinate to the first strategy of (only) braking in the ego lane. This step is subordinate because the driver perceives an unrest due to the additional steering in addition to braking, and drivers of surrounding vehicles may also be distracted by the driving maneuver of the ego vehicle. In particular, this logic should only be applied if collision avoidance is guaranteed by both strategies.

(31) A next preferred, third strategy to gain the core priority is (only) steering within the ego lane 16 of the ego vehicle 10 to avoid an obstacle, whereby this third strategy or step is preferably a strategy subordinate to the second strategy to combine braking and steering within the ego lane 16 of the ego vehicle 10. This step is subordinate to the above steps because it has been found that a steering maneuver without deceleration increases the driver's stress due to the essentially constant speed during the steering process.

(32) The priority which strategy (or step) is selected depends in particular on the prioritization between the strategies. In particular, the prioritization between combining the second strategy of braking and steering within the ego lane 16 of the ego vehicle 10 and the third strategy of (only) steering within the ego lane 16 of the ego vehicle 10 to avoid an obstacle can be based on the relative speed between the ego vehicle 10 and the obstacle. Optionally or in addition to the relative speed, the available space or distance between the ego vehicle 10 and the obstacle (e.g. third party vehicle 24) can also be relevant.

(33) Next, it may be envisaged that a strategy to gain the core priority is to full-brake in the ego lane 16 of the ego vehicle 10, whereby this step happens preferably up to the allowed emergency braking threshold, and whereby this step is preferably a strategy subordinate to steering within the ego lane 16 of the ego vehicle 10 to avoid an obstacle. Although an emergency braking maneuver increases the probability that an accident will be avoided, this step is not a common means of preventing the driver from feeling unwell on the road. For this reason, the step is subordinate to the step steering within the ego lane 16 of the ego vehicle 10 to avoid an obstacle. However, it is preferable that this step happens preferably up to the allowed emergency braking threshold, at least if the driver's safety can be guaranteed in order not to strain the driver's confidence in the ego vehicle 10.

(34) Another preferred embodiment of the invention is that a strategy to gain the core priority is to combine braking and steering towards/entering temporarily adjacent lane 12a, 12b, whereby this step is preferably a strategy subordinate to to full-brake in the ego lane 16 of the ego vehicle 10. The subordinated classification is based on the recognition that although a full-brake puts more physical strain on the driver than a lane change, but the ego vehicle 10 cannot predict the driving behavior of the surrounding vehicles, so that the driver's safety should be guaranteed at this point.

(35) According to a modified embodiment of the invention, a last strategy to gain the core priority is steering towards/entering temporarily adjacent lane 12a, 12b, whereby this step is preferably a strategy subordinate to combine braking and steering towards/entering temporarily adjacent lane 12a, 12b. One possible advantage of this sequence is that by steering to an adjacent lane 12a, 12b without braking, obstacles approaching from the front, such as uncontrolled oncoming vehicles, can be avoided faster and thus more safely than would be the case with a combination of steering and braking. The basic assumption is that a lateral accident is less risky than a frontal accident.

(36) The present invention also provides a system for providing a multi-lane scenario driving support for an ego vehicle 10 in a traffic situation, using a method according to any of the preceding features and comprising an environment sensor system 14 according to any of the preceding features and a decision device 22 according to any of the preceding features.

(37) A preferred embodiment of the invention is that the environment sensor system 14 includes a front camera. A camera has the advantage that on the one hand it is inexpensive and on the other hand it tends towards the highest possible measurement quality due to its high resolution.

(38) In order to further increase the measuring quality, it may be provided that the environment sensor system 14 includes a radar cocoon. A radar has a lower resolution than a camera, but a radar is weather-independent, because the long wavelengths are hardly disturbed by the atmosphere and therefore fog, light rain and snow can be penetrated, so that safety can be significantly increased. In addition, a radar is independent of lighting conditions, which makes day and night measurements possible, which also increases the safety. A cocoon has the advantage of 360 degrees detection.

(39) The measuring quality can also be increased if, for example the environment sensor system 14 includes a laser scanner cocoon. Laser measurements have the advantage that they are very precise and also have a very short reaction time, so that safety can be significantly increased. A cocoon has the advantage of 360 degrees detection.

(40) In order to increase the driver's safety as far as possible, it can be advantageous that measurement data from the vicinity of the vehicle are recorded as completely as possible. Therefore, a preferred embodiment of the invention is that the environment sensor system 14 includes a fully surround system.

(41) In particular, it is foreseen that the sensory units of the environment sensor system 14 may be cumulative or redundant. This increases the measurement quality and reduces the risk of wrong decisions of the decision device 22 based on missing or wrong measurement results.

REFERENCE SIGNS LIST

(42) 10 ego vehicle 12a adjacent lane 12b adjacent lane 12b adjacent lane 14 environment sensor system 16 ego lane 18 front proximity area 20 rear proximity area 22 decision device 24 third party vehicle