Method and device for the lateral guidance of a motor vehicle, in particular for assisting evasive action

0001The present invention relates to a method and a device for lateral guidance of a motor vehicle, in particular for assisting a motor vehicle in avoiding evasive action, with which a collision can be avoided. In this case, obstacles are recognized by means of an object detection sensor and information regarding their position in relation to the vehicle and at least one further property is determined and the obstacles are classified with regard to their property “can be traversed” or “cannot be traversed”. Furthermore, a trajectory is determined that describes a collision-free driving corridor away from the detected obstacles and determines steering signals according to the trajectory and outputs them to a lateral controller for evasive assistance or for automatic evasive maneuvers of the vehicle, with the classification "overpassable" or "not overpassable" based on the Object detection sensor determined height of the obstacle is performed.

State of the art

0002From the<patcit id="pcit0001" dnum="DE102006047131A1"><text>DE 10 2006 047 131 A1</text></patcit> a method for automatically controlling a vehicle is known, in which, in order to provide an avoidance trajectory of the vehicle in front of an obstacle, information about at least one area freely passable by the vehicle is taken into account.

0003the<patcit id="pcit0002" dnum="WO2013175477A1"><text>WO 2013/175477 A1</text></patcit> discloses a method for avoiding or mitigating a collision of a vehicle with an obstacle in a lane of the vehicle when the obstacle at least partially blocks a planned path of travel of the vehicle. In this case, a position and/or trajectory of the obstacle is recognized and the driving path is adapted using the position and/or the trajectory in order to avoid or mitigate the collision. An area can be recognized as passable even if an object classified as belonging to a predetermined obstacle class is recognized in the area.

0004Furthermore, both from the<patcit id="pcit0003" dnum="DE102008028222A1"><text>DE 10 2008 028222 A1</text></patcit> as well as the<patcit id="pcit0004" dnum="DE102011115353A1"><text>DE 10 2011 115353 A1</text></patcit> Methods and devices known according to the preamble of the independent claims. The object of the invention is to specify a method for transverse guidance that is further improved compared to the prior art.

0005This object is achieved by a method and a device for transverse guidance according to the independent claims.

Disclosure of Invention

0006The core of the present invention is to classify obstacles that have been detected by an object detection sensor into the classes “can be driven over” or “cannot be driven over” based on their determined height and, based on this classification, to calculate trajectories for vehicle guidance on collision-free driving corridors, with the invention providing is that recognized objects may also be driven over during trajectory planning, if they have been classified as "passable" and if no trajectory with a collision-free driving corridor can be determined , obstacles with a height below a predetermined limit value are masked out. According to the invention, this is achieved by the features of the independent claims. Advantageous developments and refinements result from the dependent claims.

0007Since detected obstacles are classified into the classes “can be driven over” and “cannot be driven over” based on their characteristic properties, it is possible that, depending on the driving situation, low-level obstacles can be driven over with a lateral guidance system. In the case of obstacles of low height, this ability to be driven over can also be assumed in case of doubt if no other collision-free travel trajectory can be determined in the alternative situation. This enables system behavior to be implemented that adapts to the respective driving situation and is based on the driver's expectations. The system is not only intended for emergency avoidance maneuvers to prevent impending collisions or to mitigate collisions with objects that can no longer be avoided, but can also be used to increase driving comfort. For example, the driver expects that a curb recognized as an obstacle during a parking maneuver represents an obstacle that can be driven over, so that according to the driver's expectation it can also be driven over at low speed during the parking maneuver. The same curb recognized as an obstacle may present itself as an obstacle at higher speeds in the event of an emergency evasive manoeuvre, which the driver expects not to be driven over, otherwise pedestrians on the adjacent sidewalk may be affected or further driving stability during the emergency maneuver may be impaired by any damage to the vehicle is at risk. Driving over this low obstacle in the form of a curb can therefore make sense at low speeds for comfort functions, but it can be avoided at high speeds and may only be driven over in an extreme emergency.

0008Provision is advantageously made here for the at least one further property of the obstacles, which is determined and provided by the object detection sensor, to be the vertical extent of the obstacle. This means the height of the upper end of the obstacle above the road surface, whereby, for example, a manhole cover or a metallic beverage can on the road surface, which each have excellent radar reflection properties, can be distinguished from a cyclist or a sign post with regard to their ability to be driven over.

0009Furthermore, it is advantageous that, in order to determine the steering signals, a measure for the criticality is also determined, which indicates how quickly and/or with what intensity the steering signals are to be implemented by the transverse controller. In particular when using the system according to the invention for emergency evasive maneuvers, the criticality indicates how quickly the steering signals have to be implemented by the actuator devices in order to defuse the momentary dangerous situation.

0010Furthermore, it is advantageous that the distance between one's own vehicle and obstacles is used to determine the criticality. In combination with this or as an alternative to this, it can be provided that the instantaneous vehicle speed is used to determine the criticality. Both the distance from the vehicle to obstacles and the instantaneous vehicle speed are variables that characterize the dangerous situation, since an impending collision can be intensified at high speeds and short distances. Furthermore, it is advantageous that the measure for the criticality is chosen to be higher, the smaller the distance between one's own vehicle and the obstacle. As an alternative to this or in addition to this, it can also be provided that the higher the instantaneous vehicle speed of one's own vehicle, the higher the degree of criticality.

0011Furthermore, it is advantageous that the predetermined limit value is selected as a function of the vehicle's own speed. This means that the higher your current vehicle speed, the lower the current, predetermined limit value must be, since even small objects that can be driven over can become a hazard at higher speeds. The lower the instantaneous vehicle speed, the more likely it is that objects with a greater vertical extent can be driven over.

0012Furthermore, it is advantageous that the predetermined limit value is selected as a function of the approach angle between one's own vehicle and the detected obstacle. Obstacles that can be driven over, such as curbs, can then be masked out if a certain driving situation occurs, so that the system does not intervene when the vehicle approaches. This is more likely to be the case, for example, at low speed and/or at a small approach angle than at high speeds or large angles between one's own travel trajectory and the object interface. Even in the case of an automatic emergency avoidance system for collision avoidance, the area for planning the emergency avoidance trajectories is limited by curbs along the roadway. Only in exceptional situations, i.e. if there is not a sufficient driving corridor available for an evasive maneuver and no collision-avoiding or collision-reducing trajectory is found, for example because these are blocked by obstacles that cannot be driven over, it can still make sense to carry out an evasive maneuver over an obstacle that can be driven over, such as a curb or to execute a traffic island if otherwise a collision eg cannot be prevented with a pedestrian. In this exceptional situation, the invention makes it possible for an emergency evasive maneuver to be carried out over an obstacle that can be driven over.

0013Of particular importance is the implementation of the method according to the invention in the form of a control element which is provided for a control unit of an adaptive distance or speed control of a motor vehicle. In this case, a program is stored on the control element, which program can be run in particular on a microprocessor or signal processor and is suitable for carrying out the method according to the invention. In this case, the invention is implemented by a program stored on the control device, so that this control element provided with the program represents the invention in the same way as the method for whose execution the program is suitable. In particular, an electrical storage medium can be used as the control element, for example a read-only memory.

0014Further features, application possibilities and advantages of the invention result from the following description of exemplary embodiments of the invention, which are illustrated in the figures of the drawing. All of the features described or illustrated form the subject matter of the invention, either alone or in any combination, regardless of how they are summarized in the patent claims or their back-reference, and regardless of their wording or Representation in the description or in the drawings.

Brief description of the drawings

0015Exemplary embodiments of the invention are explained below with reference to drawings. Show it:<dl id="dl0001" compact="compact"><dt>figure 1</dt><dd>an exemplary driving situation in which the system according to the invention can be used,</dd><dt>figure 2</dt><dd>an object scene evaluated by the sensor of the in<figref idref="f0001">figure 1</figref> described driving situation,</dd><dt>figure 3</dt><dd>a block diagram of an embodiment of the device according to the invention and</dd><dt>figure 4</dt><dd>a flowchart of an embodiment of the method according to the invention.</dd></dl>

Embodiments of the invention

0016In<figref idref="f0001">figure 1</figref> the lane 1 of a road is shown, which provides one lane in each direction of travel. Sidewalks 2 are provided on both sides of this roadway 1 and are separated from the roadway by means of a curb 3 . Curbs form an elevated step between 10 and 20 centimeters high and separate the sidewalks 2 from the roadway 1. In the direction of travel shown, the vehicle 4, which is equipped with the system according to the invention, drives on the roadway 1. This vehicle 4 is moving at an instantaneous speed v and is equipped with an object detection sensor 5 at the front of the vehicle. This object detection sensor can be an ultrasonic, lidar, video sensor or a combination of these sensor types and perceives objects in the area surrounding the vehicle 4 . The sensor viewing area 6 , which represents the detection area of ​​the object detection sensor 5 , is shown schematically for this sensor 5 . The object detection sensor 5 is able to detect the relative position of objects and their spatial extent and evaluates the object size and object position such that an occupancy map for the vehicle environment is created, for example. Such an occupancy map consists of a grid-like subdivided network, with each cell of the network being entered as to whether or not an object is detected and suspected at this relative position with respect to one's own vehicle 4 . On the basis of such an occupancy map, a lateral guidance system for vehicles, which takes over the lateral guidance task for the driver of the vehicle by actuating the steering device, can be controlled by determining a trajectory for the vehicle that describes a driving corridor that uses only unoccupied grid cells of the occupancy map for trajectory planning . Furthermore, according to the driving situation<figref idref="f0001">figure 1</figref> a vehicle 7 parked at the side of the road is shown. This parked vehicle 7 was partly parked on the adjacent sidewalk 2 and can be an object with a large height extension, for example a van or a truck. Due to the large height extent of the parked vehicle 7, the driver of the vehicle 4 cannot see that, seen in the direction of travel, a pedestrian 8 intends to cross the roadway 1 behind the parked vehicle 7. Due to the concealment situation by the parked vehicle 7, the driver of the vehicle 4 only recognizes the crossing pedestrian 8 very late, so that the collision situation can be recognized by means of the environment sensor 5, which controls the vehicle's lateral guidance. Furthermore, a beverage can 9 is shown on the roadway 1 as an example, which represents an object with a small height extent.

0017In<figref idref="f0002">figure 2</figref> is shown how the object detection sensor 5 according to the driving scene<figref idref="f0001">figure 1</figref> perceive and evaluate. The curbs 3 can be seen here, which were detected by the sensor and represent objects with a small height extent. These curbs 3 separate the sidewalks 2 from the roadway 1. Furthermore, in<figref idref="f0002">figure 2</figref> the parked vehicle 7 is shown, which can be easily detected by the object detection sensor due to its size. Furthermore, the crossing pedestrian 8 is shown, who suddenly appears behind the parked vehicle 7 in the direction of travel and forces the system according to the invention to perform an evasive maneuver. Furthermore, beverage can 9 , which is also recognized as an obstacle, is shown in the left half of lane 1 . The vehicle 4, which is approaching the parked vehicle 7 at the speed v and suddenly recognizes the crossing pedestrian 8, is forced to perform an evasive maneuver so that the pre-planned trajectory, which was originally almost straight and due to the parked vehicle 7 only with a slight lateral offset should be continued to the middle of the lane, must be modified. The system according to the invention has the task of finding a new trajectory that avoids a collision with the pedestrian 8 crossing. Finding such a trajectory turns out to be difficult in the situation shown, since the beverage can 9 was recognized as an obstacle on the roadway. By means of systems known from the prior art, the lateral vehicle control is prohibited from driving over the beverage can 9, which means that an evasive situation is difficult, since the curbs 3, which are also recognized as obstacles, must also not be driven over. Due to the configuration of the system according to the invention, it is recognized that the beverage can 9 is an obstacle with a low height extension and that it was classified as "can be driven over".

0018In contrast to this, the crossing pedestrian 8 and the vehicle 7 parked at the edge of the road were classified as "objects that cannot be driven over" due to their height extent, which is indicated by checkered hatching of the object symbols 7, 8 in<figref idref="f0002">figure 2</figref> is shown. In contrast, the left and right curbs 3, which were also recognized as an obstacle, were classified as obstacles "can be driven over", but should not be allowed to be driven over in this situation, or only in an extreme emergency, due to the higher speed v of the vehicle 4. Based on this evaluation of the object height information, the beverage can 9 is hidden and an avoidance trajectory 11 is calculated, according to which the vehicle 4 avoids the oncoming lane and thus leaves enough space to avoid a collision or endangerment of the pedestrian 8 crossing. Furthermore, in<figref idref="f0002">figure 2</figref> the driving corridor 12 drawn, which results according to the trajectory 11. The trajectory 11 forms the vehicle lane that would be left behind by a point in the center of the vehicle projected onto the road. If this trajectory 11 is traversed by the vehicle 4 , this results in a significantly wider space requirement, which is represented by the travel corridor 12 and essentially represents the left and right edges of the vehicle along the trajectory 11 . Due to the suppression of the obstacle 9, a collision situation with the suddenly appearing pedestrian 8 crossing the street could be avoided according to the invention. However, if the beverage can 9 had been another pedestrian with a correspondingly large height, the obstacle 9 would have been classified as "not traversable", so that the in<figref idref="f0001">figure 1</figref> drawn trajectory 11 would not have been possible in the manner shown. In this case, which is a modification of the in<figref idref="f0001">figure 1</figref> represents the situation shown, in which no collision-free trajectory could have been determined, it would also have been possible to hide the curb, so that a possible avoidance trajectory 11 according to this modified variant would also have used the left sidewalk 2 for avoidance.

0019In<figref idref="f0003">figure 3</figref> a block diagram of an embodiment of the device according to the invention is shown. The processing device 14 can be seen, which is accommodated as a control device in the vehicle 4 or is implemented as part of an existing control device. The processing device 14 has an input circuit 17 by means of which input signals can be supplied to the processing device 14 . Signals from an object detection sensor 5 are supplied to the processing device 14 by means of the input circuit 17 . This object detection sensor 5 is a sensor attached to the front of the vehicle 4, which can detect and evaluate obstacles in the vehicle environment using ultrasound, lidar, radar, video or a combination of the sensor technologies described. For this purpose, the relative position of the objects to the vehicle 4 and their vertical and horizontal extent are determined and transmitted to the input circuit 17 . Furthermore, the input circuit 17 is supplied with the output signal of a speed sensor 15, which detects the instantaneous vehicle speed v of the vehicle 4 and makes it available to the device according to the invention. In addition, further sensors 16 can be provided, which can optionally be added. In particular, in the case where the surroundings sensor system consists of a combination of different sensor types, the input circuit 17 can be supplied with output signals from further sensors 16 . The input circuit 17, to which the input signals of the sensors 5, 15, 16 were supplied, forwards them to a calculation device 19 by means of a data exchange device 18, which can be designed in particular as a bus system. In the calculation device 19, which can advantageously be designed as a microprocessor as a signal processor, the method according to the invention runs in the form of an algorithm and determines new output signals from the input signals, by means of which the planned travel trajectories can be traversed by downstream actuator devices. The output signals determined by the calculation device 19, which in the present case consist in particular of a steering signal 21 and a criticality signal 22, are fed via a data exchange device 18, which in turn can be designed as a bus system, to an output circuit 20, which transmits the output signals of the processing device 14 to downstream outputs actuators. The output circuit 20 forwards the output signals, which consist of the steering signal 21 and the criticality signal 22 in the present example, to the transverse controller 23, which converts signals into steering torques of the vehicle steering system. The steering signal 21 indicates when and to what extent the deflection of the steering device is intended. In this case, the criticality signal 22 specifies how quickly the steering signals 21 are to be converted into steering movements of the vehicle 4 by the transverse controller 23 and the steering device 24 .

0020In<figref idref="f0004">figure 4</figref> an exemplary sequence of the method according to the invention is shown, which is started in step S1, for example with activation of a lateral guidance support by the driver. In the subsequent step S2, object data are read in by the object detection sensor system 5 via the input circuit 17. These object data contain information regarding the relative position of the objects to the host vehicle 4, the horizontal and vertical extent of the objects, and predicted directions of movement and movement speeds. In the next step S3, an occupancy map is filled from the available object data and updated by filling the grid cells of the occupancy map with information as to whether or not an object was recognized in this grid cell. Furthermore, a certain height value is recorded and stored for the respective objects. This occupancy map results in a type of rough area map, on the basis of which a trajectory for a collision-free travel corridor of the vehicle 4 is sought and calculated in the subsequent step S4. If a trajectory was found that describes a collision-free driving corridor, the method branches to "yes" in step S5 and it is determined in a subsequent, optional step S6 with which criticality this trajectory is to be regulated by the actuator device. If in step S5 no Found a trajectory that allows a collision-free driving corridor for the vehicle 4, step S5 branches to "no" and is continued in step S9. In step S9, a limit value for object masking is specified. This limit value describes an object height, according to which all objects with a height value less than the limit value are hidden from the calculation of a trajectory for a collision-free driving corridor by changing these objects with height values ​​less than the limit value from the classification "cannot be traversed" to "can be traversed". In step S10, the objects with a lower height value than the limit value are then masked out and the method branches back to step S4, in which another attempt is made to find a trajectory with a collision-free travel corridor for vehicle 4. If a collision-free driving corridor has now been found after running through step S4 again, the method branches to "yes" in step S5, otherwise the method branches again to "no" and a new limit value is defined in step S9, which now contains a blanking of objects with a larger height value than in the previous iteration of this step. In the process, when searching for a collision-free avoidance trajectory, the process starts with a trajectory search in which only very low obstacles may be driven over. If the driving situation is so complex that no collision-free avoidance trajectory can be found, then with each iteration of the loop between step S5 and step S4, the limit value for object masking is increased to greater height values, so that step by step more objects are masked out and the probability to obtain a collision-free trajectory increases. This iteration is continued until a collision-free avoidance trajectory has been found in step S5 and the method is continued with a "yes" in that, according to step S7, the steering signals according to the calculated trajectory with a collision-free driving corridor are output to the downstream transverse controller for output to the steering actuators . In the case of the optional step S6, after which a criticality value is also determined, which indicates how quickly the steering signals are to be adjusted by the transverse controller and the steering actuator, the criticality value is also output to the transverse controller in step S7. As an alternative, the optional step S6, which provides for the determination of a criticality value, can also be inserted at a different point in the flowchart without significantly changing the inventive idea. After the steering signals and any criticality values ​​that are present have been output to the transverse controller, the algorithm ends in step S8 and can start again at step S1, for example when a new driving situation occurs in the further course of driving and a new driving trajectory has to be calculated.