DE102011106746B4 - Lane change assistance system - Google Patents

Abstract

Lane change assistance system (1) for motor vehicles for a lane change of an ego vehicle (21) from its own lane (22) to a destination lane (23) with - an environment detection unit (2) for detecting foreign vehicles (24) located in the vicinity of the ego vehicle (21) , 25a to 25c) and their respective vehicle state is set up, - a gap determination unit (5) which is set up to determine at least one sheathing gap (26a, 26b) on the destination lane (23) in dependence on the detected foreign vehicles and their respective vehicle state, a lane change judging unit which is arranged, for each determined shunting gap (26a, 26b), starting from the position of the ego vehicle (21), a plurality of target trajectories with respect to different target positions (A1 to A5, B1 to B5) within the respective shunt gap (26a , 26b) - to determine a longitudinal action space (31) of the ego vehicle (21) determining, for each target trajectory, a trajectory evaluation as a function of the determined longitudinal action space (31) taking into account predicated vehicle states at the time of reaching the target positions (A1 to A5, B1 to B6) to calculate respective target trajectories, and - an output unit (7) designed to output the trajectory scores of the target trajectories to the driver and / or to a controller of the ego vehicle, characterized in that the lane change judging unit is arranged to provide a total gap score for each determined trap trajectory Determine dependence on the trajectory ratings of the respective target trajectories of the respective Einscherlücken, and the output unit is designed to output the Trajektorienbewertungen in the form of the total gap score.

Description

The invention relates to a lane change assistance system for motor vehicles for a lane change of an ego vehicle from its own lane to a destination lane with an environment detection unit, which is set up for detecting located in the vicinity of the ego vehicle foreign vehicles and their respective vehicle state.

When coping with the driving task by the driver in road traffic more and more assistance systems are used in motor vehicles that support the driver in the accomplishment of his driving task or even independently to take over a portion of the driving task. However, as the road traffic system is characterized by a high degree of non-linearity and nondeterminism, even today autonomous driving is in an experimental state.

In particular, when changing lanes of a vehicle, the known systems quickly reach their limits, since the implementation of the lane change is a complex step, which must be composed of a variety of individual action and monitoring steps to a large complex step. Even the smallest changes in the entire scenario can fundamentally change the course of action when carrying out the lane change.

From the DE 10 2007 008 517 A1 a lane change speed assist system for vehicles is known, with which a speed specification for the desired lane change is determined. For this purpose, a lane change intent is initially recognized by the assistance system. Subsequently, the positions and speeds of the foreign vehicles are determined relative to the own vehicle, with the foreign vehicles are on the track, which intends to change the own vehicle. Subsequently, the speed specification for the lane change is calculated from these data, resulting in an optimal and safe lane change. This speed specification can then be output, for example, on an output device within the own vehicle or be made available directly as a control parameter for further assistance systems.

From the DE 101 14 187 A1 For example, there is known a method of assisting an overtaking operation wherein the distances to the vehicles in the fast lane are continuously measured and the overtaking speed is calculated in dependence on these measured distances.

From the DE 10 2005 036 714 A1 Further, a method for assisting the driver in a lane change is known, in which both vehicles on the current lane and vehicles on the destination lane are detected and assistance measures of the driver are initiated when it is detected that a lane change is imminent or carried out, which would lead to a traffic hazard. However, no assistance measures are initiated when it has been recognized that a substantial portion of the detection ranges of the sensors for the detection of the external vehicles are shadowed, so that an optimal situation image could not be determined.

From the DE 10 2004 029 369 A1 Furthermore, a lane change assistant for motor vehicles is known, which detects other road users at least on an adjacent lane with the help of environment sensors. By means of a decision device is then determined whether a lane change request of the driver is to be assumed, so that an acceleration command can be output to a speed control system. For this purpose, a window for a safe shearing to the secondary lane is determined from the data of the environment sensors and calculated an adapted acceleration strategy to this window, including the time for the start of acceleration, so that Einscheren to the secondary lane without risk is possible.

From the DE 10 2005 023 185 A1 In addition, a lane change assistance system for motor vehicles is known, which is set up by means of a monitoring device for monitoring the traffic in the apron and in the back of the own vehicle. For the decision whether a lane change is possible without danger, the temporal development of vehicle distances is calculated in advance and a recommendation for the lane change is issued when this lane change is safely possible due to the predicted distances.

From the DE 10 2006 043 149 A1 is a lateral and longitudinal guidance assistant to assist the driver in the lane change is known in which detected by means of environment sensors, the traffic environment, a Einscherlücke determined and then an optimal trajectory for the lane change is calculated taking into account the rear vehicles.

From the DE 10 2009 027 535 A1 It is further known a device for assisting a Einschervorgangs in vehicles, in which also a corresponding Einscherlücke is determined, the detection of the data is based here on a car-to-car communication. On the basis of the received data then a corresponding Einschervorgang can be calculated.

From the DE 101 34 367 A1 is a Leithilfe for a lane change known, from the distances and speeds of objects located on the target lane potential Einscherlücken determined by means of a simulation and potential Einscherlücken be signaled to the driver.

Finally, out of the DE 43 13 568 C1 a method for Leithilfe for a lane change by a motor vehicle, in which vehicles detected on the target lane, the distances monitored and if all measured distances are greater than the calculated safety distance, this is recognized as a possible lane change.

From the DE 601 23 640 T2 For example, a method and a device for predicting a travel path is known in which the travel path is calculated based on the current state of an ego vehicle and detected foreign vehicles as well as the distance already traveled, in order to be able to determine whether there is possibly a risk of collision or not.

In order to provide the driver with an optimal recommendation for a lane change even in complex driving situations, it is an object of the present invention to provide an improved lane change assistance system for motor vehicles, which gives an optimal recommendation for action especially in the presence of several gaps for the lane change.

• – eine Lückenermittlungseinheit, die zum Ermitteln mindestens einer Einscherlücke auf der Zielfahrspur in Abhängigkeit von den erfassten Fremdfahrzeugen und deren jeweiliger Fahrzeugzustand eingerichtet ist,
• – eine Spurwechseleinheit, die eingerichtet ist,
• – für jede ermittelte Einscherlücke, ausgehend von der Position des Egofahrzeuges, eine Mehrzahl von Zieltrajektorien bezüglich verschiedener Zielpositionen innerhalb der jeweiligen Einscherlücke zu bestimmen,
• – einen longitudinalen Aktionsraum des Egofahrzeuges in Abhängigkeit von fahrdynamischen Eigenschaften des Egofahrzeuges zu ermitteln, und
• – für jede Zieltrajektorie eine Trajektorienbewertung in Abhängigkeit von dem ermittelten longitudinalen Aktionsraum unter Berücksichtigung von prädizierten Fahrzeugzuständen zum Zeitpunkt des Erreichens der Zielposition der jeweiligen Zieltrajektorien zu berechnen, und
• – eine Ausgabeneinheit, die zur Ausgabe der Trajektorienbewertungen der Zieltrajektorien an den Fahrzeugführer und/oder an eine Steuereinrichtung des Egofahrzeuges ausgebildet ist.

The object is achieved with the aforementioned lane change assistance system, which according to the invention comprises:

• A gap determination unit which is set up to determine at least one lock-in gap on the destination lane in dependence on the detected foreign vehicles and their respective vehicle state,
• A lane change unit that is set up
• To determine, for each determined trapping gap, starting from the position of the ego vehicle, a plurality of target trajectories with respect to different target positions within the respective trapping gap,
• To determine a longitudinal action space of the ego vehicle as a function of the driving dynamics characteristics of the ego vehicle, and
• Calculating, for each target trajectory, a trajectory evaluation as a function of the determined longitudinal action space taking into account predicated vehicle states at the time of reaching the target position of the respective target trajectories, and
• An output unit which is designed to output the trajectory evaluations of the target trajectories to the driver and / or to a control device of the ego vehicle.

According to the invention it is thus proposed that the lane change assistance system has an environment detection unit which detects other foreign vehicles in the vicinity of the ego vehicle and their respective vehicle state. Thus, under optimal conditions with the aid of a corresponding sensor system which is connected to the surroundings detection unit via an interface, any other vehicle in the vicinity of the ego vehicle can be detected and the respective vehicle condition can be determined for each detected foreign vehicle. The data necessary for this purpose can be tapped, for example, from the internal bus system of a motor vehicle.

Such a vehicle state, which can be detected with the aid of the surroundings detection unit, can be, for example, the position of the vehicle. In addition, further vehicle states, such as the speed, acceleration or deceleration values and the relative speed to the ego vehicle can be determined as the vehicle state of a foreign vehicle.

With the aid of a gap determination unit belonging to the lane change assistance system, at least one lashing gap on the destination lane is determined from the detected foreign vehicles and their respective vehicle state. Thus, in the simplest case, the distances of the individual foreign vehicles from one another can be determined from the current position of the foreign vehicles on the destination lane, and thus those gaps which are suitable for a lane change can be selected. However, it is also conceivable that further information about the vehicle condition, such as speed, acceleration and deceleration values as well as relative speed to the ego vehicle, are taken into account when determining a suitable shearing gap.

By means of a lane change assessment unit, an evaluation is now carried out with regard to the desired lane change, so that an optimal lane change strategy can be derived as a recommended course of action. For this purpose, the lane change evaluation unit first of all determines a plurality of target trajectories with respect to different target positions within the respective lashing gap for each determined lash gap, starting from the position of the ego vehicle. This results in each Einscherlücke several target positions to which a lane change, starting from the current position of the ego vehicle, can be performed. In addition, a corresponding target trajectory for each of these target positions results, which describes the lane change path.

In parallel, the lane change evaluation unit determines the longitudinal action space of the ego vehicle as a function of its driving dynamic properties, in particular in the longitudinal direction. Such a longitudinal action space can be constructed, for example, taking into account the maximum acceleration values and the maximum deceleration values of the ego vehicle, based on its current speed. Subsequently, a score is calculated for each determined target trajectory, specifically as a function of the previously determined longitudinal action space of the ego vehicle and taking into account predicted vehicle states at the time of reaching the target position of the respective target trajectories. Thus, for example, the time of reaching the respective target position can be calculated with the aid of the longitudinal action space, as well as the predicted vehicle states of both the foreign vehicles and the ego vehicle at this time. From this, an evaluation of the respective trajectory can then be calculated, which indicates, for example, how optimal the lane change with regard to safety is to the corresponding target position of the trajectory.

For example, if the target position is close to a preceding vehicle on the target lane, it will be less optimal in terms of safety assessment if this target position is achieved with a high differential speed between the toero vehicle and the foreign vehicle. However, if the differential speed between the other vehicle and the ego vehicle when reaching the target position is negligibly small, this can be given a much higher rating.

These trajectory evaluations, which are carried out individually for each trajectory, are now output by means of an output unit to the vehicle driver and / or to a control device of the ego vehicle, so that either a corresponding recommendation for action can be given to the vehicle driver or other control systems of the motor vehicle at least the lane change partially autonomous.

For this purpose, an overall gap analysis or overall gap assessment is carried out by summarizing the respective target trajectories and their trajectory evaluations for each determined lock-in gap. Thus, for example, priority may be given to giving priority to a lock-in gap having a plurality of moderately-scored target trajectories instead of a lock-in gap that rated only a single target trajectory as good and all other target trajectories as poor.

It is particularly advantageous if the lane change evaluation unit determines the longitudinal action space of the ego vehicle on the basis of a maximum acceleration value and a maximum deceleration value as the driving dynamics characteristics of the ego vehicle. Thus, based on the instantaneous speed of the ego vehicle, maximum values for the acceleration and for the deceleration can be determined, so that, for example, trapping gaps which are outside these maximum values can not be achieved from the outset by the vehicle and thus no assessment or at least one maximum bad rating received. The longitudinal action space can thus predict the various achievable speeds of the ego vehicle over time. From this, finally, the possible achievable positions and distances can be derived over time.

Advantageously, the longitudinal action space of the ego vehicle is also determined as a function of a preceding foreign vehicle on the own lane and its vehicle state, the current traffic rules and / or the lane course of the own lane. For if there is a preceding vehicle on the own lane, which runs at a slower speed than the ego vehicle, so the longitudinal action space of the ego vehicle is also limited by this other vehicle. The same applies to the lane course of the own lane, which must be left, for example, in an acceleration strip within a fixed path. But even the currently valid traffic rules in the vicinity of the vehicle can cause the longitudinal action space must be reduced accordingly, as actions outside the permissible traffic rules to be prevented from the outset.

By means of the thus constructed longitudinal action space of the ego vehicle, the time of reaching the target position as well as the vehicle state existing at that time, for example the speed, can then be determined for each target trajectory. Taking into account the vehicle state of the ego vehicle at the respective target position at the time of reaching and the vehicle states of the respective foreign vehicle at this time, the target trajectory can then be correspondingly evaluated.

From the vehicle states of the foreign vehicles, such as speed information and position information of the respective foreign vehicles as vehicle states, then a predicted vehicle state can be determined over time, which predicts the vehicle state at the time of reaching the target positions. With Help of these predicted vehicle states of the foreign vehicles and the vehicle state of the ego vehicle at the time of reaching the target position can then evaluate the corresponding trajectory for this target position. From these evaluations, the optimal target trajectory and thus the optimal target position for the shingling gaps can be determined.

Advantageously, the trajectory evaluations are output by the output unit in the form of an action recommendation to the driver of the ego vehicle, so that the latter receives the optimal lane change strategy and can then follow this, for example to obtain an optimally reliable result.

Moreover, it is particularly advantageous if the lane change assistance system is set up in such a way that it continuously carries out a trajectory evaluation for gap analysis so that, for example, the driver can permanently be shown the optimum shedding gap and shearing position. For this purpose, the necessary information is permanently recorded and provided by the systems so that, for example, the changing longitudinal action space and the change in the vehicle states of the foreign vehicles and the ego vehicle can permanently react to the changing situations. Thus, the optimal lane change strategy can be determined at any time.

The invention will be described by way of example with reference to the accompanying drawings. Show it:

1 - schematic structure of the lane change assistance system according to the invention,

2 Schematic representation of an embodiment with two Einscherlücken,

3 Schematic representation of the longitudinal action space,

4a to 4c - Exemplary trajectory ratings.

1 schematically shows the structure of the lane change assistance system according to the invention 1 , The lane change assistance system 1 has an environment detection unit 2 on, which is set up for detecting located in the vicinity of the ego vehicle foreign vehicles and their respective vehicle state. The detection of the vehicles and their vehicle states, for example, using a sensor 3 carried out with the environment detection unit 2 contacted via an interface. However, the necessary data can also be provided by the surroundings detection unit 2 from the bus system 4 of the vehicle with which the surroundings detection unit 2 is connected via an interface, be tapped.

The so from the environment detection unit 2 recorded data, ie the foreign vehicles and their respective vehicle state are then for further calculation to a gap determination unit 5 which is arranged to determine at least one Einscherlücke on the target lane, which the ego vehicle intends to change depending on the detected foreign vehicles with their respective vehicle state. As the vehicle state in the sense of the present invention, information representing the state of the vehicle, such as the position of the vehicle, the speed of the vehicle, acceleration or deceleration values of the vehicle as well as the relative speed to the ego vehicle are considered. With the help of the position of the foreign vehicles as information about the vehicle condition can thereby be determined by the gap determination unit 5 Determine the corresponding lashing gaps on the target lane, which can be used to cut into this target lane.

Both the data of the environment detection unit 2 as well as the data of the gap determination unit 5 are sent to a lane change assessment unit 6 forwarded, then makes a corresponding assessment of the lane change. This is done first for each in the gap determination unit 5 determined Einscherlücke determined corresponding Zieltrajektorien with respect to a plurality of target positions within a Einscherlücke to which the ego vehicle can switch to the destination lane. Parallel to this, the longitudinal action space of the ego vehicle is set up as a function of the driving dynamics of the ego vehicle. The lane change evaluation unit 6 is now set up in such a way that it evaluates the individual target trajectories as a function of the longitudinal action space taking into account the vehicle states as they would at the time the respective target position of the trajectory was reached, so that a trajectory evaluation is present for each target trajectory at the end.

The evaluation is carried out in such a way that, considering the circumstances, it indicates how sensible or even safe the lane change is on this target trajectory.

The trajectory ratings can now be on one output unit 7 be issued to the driver, for example, to indicate a corresponding action recommendation for the lane change. However, the output can also be sent to a controller 8th of the vehicle are spent, thus, for example, the lane change is carried out autonomously by the ego vehicle itself or at least supports the Fahrreugführer when changing lanes accordingly.

Based on the example of the 2 Let us briefly explain the operation of the lane change assistance system of the present invention.

An ego vehicle 21 is on his own lane 22 and intends to get on the target lane 23 cross over. Both on the own lane 22 as well as on the destination lane 23 are foreign vehicles 24 . 25a to 25c , The foreign vehicle 24 is on the same lane 22 like the ego vehicle 21 while the foreign vehicles 25a to 25c on the destination lane 23 on the way.

First, the possible Einscherlücken 26a and 26b determined. The shedding gap 26a is located between the foreign vehicles 25a and 25b while the hernia gap 26b between the foreign vehicles 25b and 25c is determined. Within each of these Einscherlücken 26a and 26b can be determined target positions A1 to A5, B1 to B5, which can be regarded as a destination in a lane change. For each of these target positions A1 to A5, B1 to B5 of the respective Einscherlücken 26a . 26b Trajectories can now be determined on which the ego vehicle 21 from his own lane 22 on the destination lane 23 to the destination of the corresponding destination position. The representation of the gaps 26a . 26b and their respective target positions A1 to A5, B1 to B5 is also used as a gap band 27 designated.

By means of the below further explained longitudinal action space of the ego vehicle 21 Now it is possible to carry out a trajectory evaluation for each trajectory on the target position, whereby the vehicle states of the foreign vehicles 25a to 25c be predicted at the time of reaching the respective target position and thus taken into account.

Thus results for each target position A1 to A5, B1 to B5 and their respective trajectory, starting from the position of the ego vehicle 21 , a corresponding Trajektorienbewertung, which then presented as a recommendation to the driver or a control unit for autonomous control of the ego vehicle 21 can be supplied.

3 shows in the two diagrams the longitudinal action space of the ego vehicle, which ultimately describes the possible action space using the longitudinal dynamic capabilities of the vehicle. In 3 Once in the upper diagram, the possible speeds over time are shown as a longitudinal action space, while the lower diagram shows the possible achievable positions over time.

One way to build up the longitudinal action space of the ego vehicle is to combine the maximum acceleration values and the maximum deceleration values of the ego vehicle. For this purpose, there is then once the downward slope 32 for the maximum delay value and the upward facing limit 33 for the maximum acceleration value, based on the current speed of the ego vehicle (25 m / s in this example). The diamond pattern of the longitudinal action space 31 This is based on the assumption that for a certain period of time initially a maximum acceleration value is assumed, and then for a further period a maximum deceleration value. Thus, starting from the initial speed of 25 m / s, the vehicle first accelerates for a maximum of 0.2 seconds and then decelerates with the maximum deceleration, resulting in the illustrated diamond pattern, depending on the combination.

In addition, the longitudinal action space upwards is also limited, which is partly due to the longitudinal dynamic properties of the vehicle, but on the other hand caused by other circumstances. Such is how in 2 shown, the own lane not free, so that the longitudinal action space is limited by a foreign vehicle on the own lane. But also the geometry of the lane, on which the own vehicle currently drives, can the longitudinal action space 31 limit as well as limiting traffic rules at the current position of the ego vehicle.

For this reason, the longitudinal action space is in 3 limited to the top and flattening backwards, so that target positions outside the longitudinal action space 31 can not be reached.

The lower diagram shows the one from the longitudinal action space 31 deductible positions over time that can be achieved by the ego vehicle.

From the longitudinal action space 31 For each target trajectory, assuming the maximum acceleration and deceleration values of the ego vehicle, it is then possible to determine the times when the target position of the respective target trajectory is reached and which vehicle state the ego vehicle then has (position and speed). Taking into account predicated vehicle states of the foreign vehicles at the time of reaching the target positions of the respective target trajectories can then be an assessment of how good or how bad the corresponding trajectory for the lane change, for example, in terms of safety.

As a result, there is then a rating for each target trajectory, which can then be plotted over the entire gap band, so that a complete assessment of the overall situation can be derived. This is exemplary in the 4a) to 4c) shown.

In 4a) In the upper diagram, the evaluation of the individual target trajectories from 0 to 1 is removed, as this would result from the situation below. The ego vehicle drives at a speed of 25 m / s, while a foreign vehicle travels on the target lane at a slower speed of 15 m / s. The ego vehicle now has two options, either scissors behind the other vehicle or accelerate and sheared in front of the other vehicle. With the above lane change assistance system, an evaluation of the individual trajectories for the lane change now takes place, the highest rating being obtained at the current position of the ego vehicle. At the level of the other vehicle, the rating is 0, as expected, since safe shearing would not be possible without an accident. Cabling in front of the target vehicle was rated at a maximum of 0.6, whereby, of course, the lane course of one's own lane was also taken into account here.

The maximum evaluation of the target trajectory can be output as a recommendation for action to the driver, so that a lane change recommendation behind the other vehicle is recommended.

4b) shows an example in which the ego vehicle is almost at the level of the other vehicle on the target lane and moves at a much higher speed than the other vehicle. A lane change at the current position was correctly evaluated by the system according to the invention as 0, as this would lead to an accident. However, the system rates the shearing in front of the other vehicle as the best option, so that a corresponding action recommendation can be issued here as well.

4c) shows an example in which the other vehicle on the target lane moves at a higher speed than the own vehicle on the own lane. A shearing in front of the other vehicle on the target lane was therefore rated 0 by the system according to the invention and thus not recommended. Taking into account all the circumstances mentioned above, a lane change at the current position was rated as the best action.

By means of this representation, the driver can be shown, by way of example, an overall gap evaluation, from which he can deduce which lashing gap and in particular which target position within the lashing gap is the best way to perform a safe lane change. With the aid of the longitudinal action space, the possible state space of the ego vehicle can be represented in the longitudinal direction.

Claims (5)

Lane change assistance system ( 1 ) for motor vehicles for a lane change of an ego vehicle ( 21 ) from its own lane ( 22 ) on a target lane ( 23 ) with - an environment detection unit ( 2 ) used for detecting in the vicinity of the ego vehicle ( 21 ) ( 24 . 25a to 25c ) and their respective vehicle state is set up, - a gap determination unit ( 5 ) used to determine at least one shedding gap ( 26a . 26b ) on the destination lane ( 23 ) is set up as a function of the detected foreign vehicles and their respective vehicle state, - a lane change assessment unit which is set up, - for each determined lashing gap ( 26a . 26b ), starting from the position of the ego vehicle ( 21 ), a plurality of target trajectories with respect to different target positions (A1 to A5, B1 to B5) within the respective shedding gap ( 26a . 26b ), - a longitudinal action space ( 31 ) of the ego vehicle ( 21 ) depending on the driving dynamics of the ego vehicle ( 21 ) and, for each target trajectory, a trajectory evaluation as a function of the determined longitudinal action space ( 31 taking into account predicted vehicle states at the time of reaching the target positions (A1 to A5, B1 to B6) of the respective target trajectories, and - an output unit ( 7 ), which is designed for outputting the trajectory evaluations of the target trajectories to the vehicle driver and / or to a control device of the ego vehicle, characterized in that the lane change evaluation unit is set up, for each determined shedding gap, an overall gap evaluation depending on the trajectory evaluations of the respective target trajectories of the respective shingling gaps determine, and the output unit is designed to output the Trajektorienbewertungen in the form of the overall gap assessment. Lane change assistance system ( 1 ) according to claim 1, characterized in that the lane change evaluation unit ( 6 ), the longitudinal action space ( 31 ) of the ego vehicle ( 21 ) on the basis of a maximum acceleration value and a maximum deceleration value as the driving dynamics characteristics of the ego vehicle ( 21 ) to investigate. Lane change assistance system ( 1 ) according to claim 1 or 2, characterized in that the lane change evaluation unit ( 6 ), the longitudinal action space ( 31 ) of the ego vehicle ( 21 ) as a function of a preceding vehicle ( 24 ) on the own lane ( 22 ) and its vehicle state, the current traffic rules and / or the lane course of the own lane ( 22 ) to investigate. Lane change assistance system ( 1 ) according to one of the preceding claims, characterized in that the lane change assessment unit ( 6 ) is set up for each target trajectory to determine the time of reaching the target position (A1 to A5, B1 to B5) of the respective target trajectory by the ego vehicle and the vehicle state of the ego vehicle at the target position in dependence on the longitudinal action space and the target trajectories taking into account Assess vehicle state of the ego vehicle at the respective target position at the time determined. Lane change assistance system ( 1 ) according to one of the preceding claims, characterized in that the output unit ( 7 ), the trajectory evaluations in the form of a recommendation for action to the driver of the ego vehicle ( 21 ).

Images