METHOD FOR OPERATING A VEHICLE, PARKING ASSISTANCE SYSTEM AND VEHICLE

Description

[0001] The present invention relates to a method for operating a vehicle, a parking assistance system and a vehicle.

[0002] Parking assistance systems for vehicles are known that can be trained by a user in a training mode to follow a specific trajectory with the vehicle, for example, from a yard entrance to a garage. For this purpose, the trajectory manually driven by the user is recorded. During subsequent following, the parking assistance system automatically executes the control actions performed by the user during training, causing the vehicle to follow the trajectory. DE 10 2018 117718 A1 describes a method for operating a vehicle using a parking assistance system that is configured in a following mode to drive the vehicle along a trained trajectory or to initiate driving.

[0003] DE 10 2017 002 731 A1 describes a method for operating a motor vehicle. The method comprises manually guiding the motor vehicle along a first trajectory during a first parking process, in particular a parking process. The first trajectory is acquired during the first parking process. Furthermore, a second trajectory is determined by applying a smoothing method to the first trajectory and / or by adapting the first trajectory to environmental data acquired during the first parking process, in particular to obstacles and / or obstacle-free areas determined from the acquired environmental data.

[0004] For technical reasons, a very high or even maximum steering angle, which the user can easily achieve when training the trajectory, may not be achievable in tracking mode. Therefore, the parking assistance system is designed to interrupt trajectory training if the user exceeds an upper limit for the steering angle. The trajectory is then not saved and is not available for tracking. If the parking assistance system nevertheless trains and saves the trajectory, the problem may arise that the vehicle deviates from the trained trajectory while following it. This could lead to the abort of the tracking maneuver at best and, at worst, to the vehicle colliding with an obstacle.

[0005] Against this background, one object of the present invention is to improve the operation of a vehicle.

[0006] According to a first aspect, a method for operating a vehicle by means of a parking assistance system is proposed, which is configured in a follow-up mode for driving the vehicle along a trained trajectory or for initiating the driving of the vehicle. In a first step a), the trained trajectory, which is acquired in a training mode in which a manually driven trajectory is recorded as the trained trajectory, is obtained. "Acquired" here means, in particular, that at least raw measurement data from a sensory acquisition of the trajectory is provided. Preferably, the trained trajectory is previously acquired by sensory means, in particular with the aid of at least one vehicle sensor, such as an ultrasonic sensor and / or camera sensor and / or GPS sensor and / or radar sensor and / or lidar sensor and / or odometry sensor.The trained trajectory comprises at least one sharply curved section whose radius of curvature is smaller than a radius that can be achieved with the vehicle using the largest permissible steering angle specified in the tracking mode. In a second step b), a starting point and an end point of the sharply curved section in the trained trajectory are determined. The determination also includes setting and / or recording. In a third step c), a substitute trajectory is determined that connects a substitute starting point with a substitute end point. The substitute starting point and the substitute end point each lie on the trained trajectory, and the sharply curved section lies between the substitute starting point and the substitute end point. A maximum curvature of the substitute trajectory has a radius of curvature that can be achieved with the largest permissible steering angle specified in the tracking mode or a smaller steering angle.

[0007] This method has the advantage that, on the one hand, trajectories that include very tight curves can be trained in a training mode, and, on the other hand, the trained trajectory can be successfully followed in the follow-up mode without the vehicle straying too far from the trained trajectory. This method is preferably used where the traffic situation does not necessarily require the user to apply the maximum steering angle when training the trajectory. Steering angle, or steering angle, is understood, for example, as an angle between a wheel rotation axis of a steering wheel and a transverse direction of the vehicle, or between a wheel rolling direction and a longitudinal direction of the vehicle.

[0008] The parking assistance system, which can also be referred to as a driver assistance system, is designed specifically for semi-autonomous or fully autonomous driving of the vehicle.

[0009] Semi-autonomous driving, for example, means that the parking assistance system controls a steering device and / or an automatic gearshift. Fully autonomous driving, for example, means that the parking assistance system also controls a drive device and a braking device. The parking assistance system can be implemented using hardware and / or software. In a hardware implementation, the parking assistance system can be designed, for example, as a computer or a microprocessor. In a software implementation, the parking assistance system can be designed as a computer program product, as a function, as a routine, as part of a program code, or as an executable object. In particular, the parking assistance system can be designed as part of a higher-level control system of the vehicle, such as an ECU (Engine Control Unit).

[0010] The vehicle is, for example, a passenger car or a truck. The vehicle comprises, in particular, a number of sensor units configured to detect the driving state of the vehicle and the surroundings of the vehicle. Examples of such sensor units of the vehicle include image recording devices such as a camera, a radar or a lidar, ultrasonic sensors, a positioning sensor, wheel angle sensors, and / or wheel speed sensors. The sensor units are each configured to output a sensor signal, for example, to the parking assistance system, which performs semi-autonomous or fully autonomous driving depending on the detected sensor signals.

[0011] To train a trajectory, the parking assistance system is, for example, placed into a training mode by the user. In the training mode, the parking assistance system preferably records all control inputs from the user, such as a steering angle, a gear engaged, a distance traveled, and the like. The corresponding values ​​are preferably stored in relation to one another, so that, for example, the steering angle is available as a function of the wheel revolutions. This clearly defines the trained trajectory, at least during slip-free driving. In addition to this data, which is also referred to as odometry data, sensor signals from the vehicle's environmental sensors are preferably also recorded during training. This can facilitate orientation of the parking assistance system, particularly in the follow-up mode.

[0012] A special feature here is that the trajectory is recorded even if a higher steering angle is used than can be achieved in tracking mode. Very high steering angles, such as a maximum steering angle, require very high torque. The user can easily achieve this torque using the steering wheel and the servo motor, but this places a very high load on an actuator used to control the steering angle in autonomous driving. To minimize the risk of rapid wear or premature failure of the actuator, the maximum steering angle achievable in autonomous driving mode can therefore be reduced; for example, it can be 80% or even 90% of the design maximum steering angle. The maximum achievable steering angle in tracking mode can also be referred to as the limit steering angle.

[0013] The larger the steering angle, the greater the curvature of the resulting trajectory and the smaller the radius of curvature of a circle of curvature or osculating circle. Accordingly, a trajectory that can be driven in tracking mode has a maximum curvature limit and a minimum curvature limit. A curvature greater than the limit and a curvature radius smaller than the limit are therefore not achievable.

[0014] The trained trajectory can also comprise several subsections whose curvature is greater than the limit curvature. The respective trajectory, for example, the trained trajectory and the substitute trajectory, refers, for example, to a reference point, in particular the center point, center of gravity, or center point of the vehicle's rear axle. Alternatively or additionally, trajectories for one or more wheels of the vehicle can be recorded and / or determined and / or obtained, and / or a trajectory for one or more corner points of a vehicle body.

[0015] The starting point and end point of the sharply curved section can be determined during trajectory recording or training. For example, a data point in the odometry data corresponding to a time at which the user exceeds or falls below the limit steering angle is marked accordingly. Alternatively, the curvature of the trajectory can be analyzed at each point of the trajectory only after the tracking mode has been activated and the corresponding trajectory has been selected, and connected sections whose curvature exceeds the limit curvature can be marked accordingly. The first point of such a section is then the starting point, and the last point of the section is the end point.

[0016] Subsequently, a replacement trajectory is determined for the highly curved section, the maximum curvature of which does not exceed the limit curvature. The replacement start point and end point of the replacement trajectory both lie on the trained trajectory and include the highly curved section. This ensures that the replacement trajectory deviates as little as possible from the trained trajectory and that the vehicle returns to the trained trajectory after following the replacement trajectory.

[0017] For complex trained trajectories that have several highly curved sections in close succession, such as an S-curve, it can be provided that a replacement trajectory is determined for individual ones of the highly curved sections, wherein a replacement start point and / or a replacement end point of the respective replacement trajectory does not lie on the trained trajectory. The several replacement trajectories determined in this way are then combined to form an overall replacement trajectory whose start point and end point lie on the trained trajectory. The overall replacement trajectory does not necessarily comprise the entire course of the several replacement trajectories; rather, they can be partially shortened. For example, transition points are determined for this purpose at which a transition takes place from a first replacement trajectory to a second, subsequent replacement trajectory.Such a transition point can be located between the substitute starting point and the substitute end point of a respective substitute trajectory.

[0018] In embodiments, the determined substitute trajectory is stored, and in particular, information regarding whether the determined substitute trajectory was successfully retraced can be stored. Then, the next time the user wishes to retrace the trained trajectory, the stored trajectory can be used. This can reduce the required computational effort.

[0019] According to one embodiment of the method, the replacement trajectory runs at least partially within the strongly curved sub-section and / or the replacement trajectory has a sub-section whose curvature is negative compared to the curvature of the strongly curved sub-section and / or the replacement trajectory comprises a number of sub-sections where the direction of travel changes.

[0020] The fact that the replacement trajectory runs within the strongly curved section means, for example, that the circle of curvature or osculating circle, which is a circle drawn with the radius of curvature and adjacent to the strongly curved section, is intersected.

[0021] Negative curvature means that the curve runs in the opposite direction while traveling in the same direction, corresponding to a change in direction from a left turn to a right turn, or vice versa. One can also speak of a convex or concave curvature. One can also say that the circle of curvature lies on the other side of the trajectory.

[0022] A change of direction section means that the vehicle changes direction in this section, i.e. changes from forward to reverse or vice versa.

[0023] According to a further embodiment of the method, the substitute starting point lies in front of the starting point of the strongly curved sub-section on the trained trajectory with respect to a direction of travel of the vehicle and / or the substitute end point lies behind the end point of the strongly curved sub-section on the trained trajectory with respect to a direction of travel of the vehicle.

[0024] This ensures that the replacement trajectory completely encloses the strongly curved section and, for example, that an end point of the trained trajectory can be safely approached.

[0025] According to a further embodiment of the method, the replacement trajectory is shorter than the strongly curved section.

[0026] This is especially the case if the replacement trajectory runs largely or entirely within the strongly curved section.

[0027] According to a further embodiment of the method, a maximum offset between the trained trajectory and the replacement trajectory is smaller than a predetermined limit value.

[0028] The offset can also be referred to as distance. The maximum offset at a given position in the replacement trajectory is, in particular, the smallest distance to a point along the trained trajectory. The specified threshold is, for example, 1 m, 2 m, 3 m, 4 m, or up to 5 m. Different thresholds for the maximum offset can be specified for different environments, such as an underground garage, a parking garage, a parking area in a courtyard, or along a street.

[0029] Preferably, the vehicle's geometry is also taken into account, so that, for example, a different orientation of the vehicle also contributes to the offset. The vehicle's geometry can be approximated, for example, by the area covered by the vehicle. Various dimensions can be introduced as the offset. For example, all points that lie outside the area swept by the vehicle in training mode can be specified as the offset. Alternatively, an average or an RMS value of the smallest distance between these points and the swept area can be defined as the offset.

[0030] This embodiment ensures that the vehicle does not deviate too far from the trained trajectory when following the replacement trajectory.

[0031] According to a further embodiment of the method, the replacement trajectory comprises at least one change of direction subsection whose curvature is smaller than the curvature of the replacement trajectory at a subsection starting point of the change of direction subsection, preferably zero, preferably negative.

[0032] One could also say that counter-steering occurs during the direction change section. This allows a deviation in the vehicle's orientation from the vehicle's orientation on the trained trajectory to be corrected particularly efficiently.

[0033] According to a further embodiment of the method, an opening angle between a first tangent to the substitute trajectory at the subsection starting point of the respective change of direction subsection of the number and a second tangent to the trained trajectory at a point corresponding to the subsection starting point is greater than or equal to a predetermined limit value.

[0034] The corresponding point, for example, is the point on the trained trajectory that has the shortest distance to the subsection starting point. The specified threshold for the aperture angle is, for example, 15°, 20°, 25°, 30°, 35°, 40°, or up to 45°. Different thresholds can be specified for different environments, such as an underground garage, a parking garage, a parking area in a courtyard, or along a street.

[0035] According to a further embodiment of the method, the replacement trajectory comprises a change of direction section after each predetermined distance along the replacement trajectory.

[0036] The specified route can be defined as a distance traveled. However, the specified route can also refer to a change of direction; for example, a change of direction section can be provided after each 45° turn or a 90° turn.

[0037] According to a further embodiment of the method, the replacement trajectory has a negative curvature with respect to the strongly curved subsection in an initial section which directly adjoins the replacement starting point and / or in an end section which directly adjoins the replacement end point.

[0038] According to a further embodiment of the method, the replacement start point, the replacement end point and the replacement trajectory are determined by an iterative optimization method.

[0039] Iterative optimization methods are particularly suitable for considering various predefined limits, such as a maximum permissible offset and / or opening angle. Furthermore, iterative optimization methods can account for dynamically changing situations, such as moving objects or obstacles.

[0040] In embodiments, environmental sensor data provided by environmental sensors of the vehicle and / or by external environmental sensors are taken into account in the determination.

[0041] This embodiment is advantageous because temporal changes in the traffic situation, such as mobile obstacles, are taken into account when determining the replacement trajectory.

[0042] According to a further embodiment of the method, a localization of the vehicle is carried out at a localization position on the trained trajectory which lies before the starting point of the strongly curved subsection.

[0043] Localization in this context refers to determining the vehicle's location. The location can be specified, for example, by coordinates on a spherical surface that describes the Earth and can be determined, in particular, using a satellite-based system such as GPS. However, the location can also be determined relative to stationary objects, such as buildings or road markings. Localization can be performed by an internal device based on environmental sensor data, or by an external device, such as an external camera or the like.

[0044] In embodiments, the position of the vehicle can be determined using odometry while following the trained trajectory. Alternatively, the vehicle's localization can be repeated periodically or on an ad hoc basis. In this way, drifting of the actual location, for example, due to inaccuracies in the initial localization and / or during odometry, can be avoided.

[0045] According to one embodiment of the method, the largest permissible steering angle specified in the follow-up mode is different for left and right.

[0046] Due to design differences in a wheel suspension and the like, differences in the steering behavior of the vehicle can arise, which can be taken into account here.

[0047] According to a second aspect, a computer program product is proposed which comprises instructions which, when the program is executed by a computer, cause the computer to carry out the method described above.

[0048] A computer program product, such as a computer program means, can be provided or delivered, for example, as a storage medium, such as a memory card, USB stick, CD-ROM, DVD, or in the form of a downloadable file from a server in a network. This can be done, for example, in a wireless communications network by transmitting a corresponding file with the computer program product or the computer program means.

[0049] According to a third aspect, a parking assistance system for a vehicle is proposed, which, in a tracking mode, is configured to drive the vehicle along a trained trajectory. The parking assistance system comprises a computing unit for obtaining the trained trajectory, which is a trajectory manually driven during a training mode, wherein the trained trajectory comprises at least one sharply curved subsection whose radius of curvature is smaller than can be achieved with the vehicle using the largest permissible steering angle specified in the tracking mode. Furthermore, the parking assistance system comprises a determination unit for determining a starting point and an end point of the sharply curved subsection in the trained trajectory, and a determination unit for determining a substitute trajectory that connects a substitute starting point with a substitute end point.The substitute start point and the substitute end point lie on the trained trajectory and the strongly curved section lies between the substitute start point and the substitute end point, and a maximum curvature of the substitute trajectory has a radius of curvature that can be achieved with the largest permissible or a smaller steering angle specified in the tracking mode.

[0050] This parking assistance system has the same advantages as explained for the method of the first aspect. The embodiments and features proposed for the described method apply accordingly to the proposed parking assistance system. The parking assistance system is operated, in particular, using the method according to the first aspect.

[0051] The respective unit, for example, the computing unit, the determination unit, and / or the detection unit, as well as the parking assistance system, can be implemented in hardware and / or software. In a hardware implementation, the unit can be configured, for example, as a computer or a microprocessor. In a software implementation, the unit can be configured as a computer program product, as a function, as a routine, as part of a program code, or as an executable object.

[0052] According to a fourth aspect, a vehicle with a parking assistance system according to the third aspect is proposed.

[0053] The vehicle is preferably a passenger car or a truck.

[0054] Further advantageous embodiments and aspects of the invention are the subject of the dependent claims and the exemplary embodiments of the invention described below. The invention will be explained in more detail below using preferred embodiments with reference to the accompanying figures. Fig. 1 shows a schematic view of a vehicle from a bird's eye view; Fig. 2 schematically shows a first example of a replacement trajectory; Fig. 3 schematically shows a second example of a replacement trajectory; Fig. 4 schematically shows a third example of a replacement trajectory; Fig. 5 schematically shows a fourth example of a replacement trajectory; Fig. 6 schematically shows a trajectory with circles of curvature; Fig. 7 shows a schematic block diagram of an embodiment of a method for operating a parking assistance system; and Fig. 8 shows a schematic block diagram of an example of a parking assistance system.

[0055] In the figures, identical or functionally equivalent elements have been given the same reference numerals unless otherwise stated.

[0056] Fig. 1 shows a schematic view of a vehicle 100 from a bird's eye view. The vehicle 100 is, for example, a car that is arranged in an environment 200. The car 100 has a parking assistance system 110, which is designed, for example, as a control unit. In addition, a plurality of sensor devices 120, 130 are arranged on the car 100, which are, for example, optical sensors 120 and ultrasonic sensors 130. The optical sensors 120 include, for example, visual cameras, a RADAR ( radio detection and ranging ) and / or a LIDAR ( light detection and ranging ) .The optical sensors 120 can each capture an image of a respective area of ​​the environment 200 of the car 100 and output it as an optical sensor signal. The ultrasonic sensors 130 are configured to detect a distance to objects located in the environment 200 and to output a corresponding sensor signal. Using the sensor signals captured by the sensors 120, 130, the parking assistance system 110 is capable of driving the car 100 semi-autonomously or fully autonomously. In addition to the Fig. 1 In addition to the optical sensors 120 and ultrasonic sensors 130 shown, the vehicle 100 may be provided with various additional sensor devices 120, 130. Examples of these include a microphone, an acceleration sensor, an antenna with a coupled receiver for receiving electromagnetically transmittable data signals, and the like. The vehicle 100 preferably further comprises a localization unit (not shown) configured to detect a position of the vehicle 100 in the environment 200 and output it to the parking assistance system 110.

[0057] Parking assistance system 110 is for driving the vehicle 100 along a trained trajectory TR (see Fig. 2 - 6 ) in a follow-up mode. The parking assistance system 110 comprises, for example, a computing unit 111 (see Fig. 8 ), a destination unit 112 (see Fig. 8 ) and an investigative unit 113 (see Fig. 8 ).

[0058] Fig. 2 shows a schematic representation of a first example of a replacement trajectory ETR, which was determined from a trained trajectory TR. The trained trajectory TR begins at an initial position AP, on which the vehicle 100 is indicated with the intended direction of travel DIR, and ends at an end position EP. The trained trajectory TR has a strongly curved section TA, which begins at a starting point A and extends to an end point E, thereby describing a 90° curve. The steering angle required to follow the strongly curved section TA with the vehicle 100 is higher than a value in the follow-up mode of the parking assistance system 110 (see Fig. 1 or 8) maximum permissible steering angle. Therefore, the parking assistance system 110 determines a substitute trajectory ETR that connects a substitute starting point EA with a substitute end point EE and whose curvature is achieved with a steering angle that can be reached in the follow-up mode. The substitute starting point EA lies on the trained trajectory TR before the starting point A of the strongly curved sub-section TA. The substitute end point EE lies on the trained trajectory TR after the end point E of the strongly curved sub-section TA. The substitute trajectory ETR thus intersects the curve defined by the strongly curved sub-section TA. By the parking assistance system 110 following the trained trajectory TR using the substitute trajectory ETR, the vehicle 100 moves from the initial position AP to the final position EP without exceeding the limit steering angle and without a maximum deviation DT from the trained trajectory TR exceeding an upper limit value.The upper limit for the maximum offset DT is preferably predetermined and can be different for different trained trajectories TR or for different sections along the trained trajectory TR. Furthermore, the upper limit for the maximum offset DT can be adjusted depending on the nature of the environment 200 (see . Fig. 1 ), in particular, obstacles or objects present in the environment 200. It should also be noted that the maximum permissible offset DT does not have to be exhausted. Preferably, the replacement trajectory ETR is determined such that the maximum offset is minimal at every point.

[0059] Fig. 3 schematically shows a second example of a replacement trajectory ETR for a trained trajectory TR, which in this case has a strongly curved sub-section TA that describes a 180° curve. The curvature of the strongly curved sub-section TA is greater than the maximum achievable curvature in the tracking mode. The replacement trajectory ETR comprises three sub-sections: a first sub-section ETA1, a second sub-section ETA2, and a direction change sub-section RWTA. The curvature of the replacement trajectory ETR is less than or equal to the maximum permissible curvature at every point. The first sub-section ETA1 connects the replacement starting point EA, which lies on the trained trajectory TR, with a first reversal point U1, at which the change of direction is carried out. The direction change sub-section RWTA connects the first reversal point U1 with a second reversal point U2, at which a further change of direction is carried out.The direction of travel DIR of vehicle 100 on the direction-change subsection RWTA is thus reversed to the direction of travel DIR of vehicle 100 at the other points of the replacement trajectory ETR. The second subsection ETA2 connects the second reversal point U2 with the replacement end point EE, which again lies on the trained trajectory TR.

[0060] It can be seen that the replacement trajectory ETR runs partly inside and partly outside the strongly curved section TA. The direction change section RWTA was determined such that a maximum permissible offset DT from the trained trajectory TR is not exceeded. The maximum permissible offset DT is, for example, 0.3 m, 0.5 m, 1 m, 2 m, 3 m, 4 m, or up to 5 m, where the offset DT is defined, for example, as a distance between the two trajectories. By reversing the vehicle 100 in the direction change section RWTA, it is ensured, for example, that the replacement trajectory ETR does not continue to the right beyond the trained trajectory TR, which could be blocked, for example, by a structural obstacle (not shown).

[0061] It should be noted that the replacement trajectory ETR can also be determined such that it runs entirely within the highly curved subsection TA, for example, by shifting the replacement starting point EA further toward the starting position AP and the replacement end point EE further toward the end position EP, and / or by inserting multiple change-of-direction subsections RWTA, each of which lies within the highly curved subsection TA. In this case, the replacement trajectory ETR would no longer intersect or cross the trained trajectory TR.

[0062] Fig. 4 schematically shows a third example of a replacement trajectory ETR for a trained trajectory TR, which in this case has a strongly curved section TA that describes a 180° curve. The curvature of the strongly curved section TA is higher than the maximum curvature achievable in the tracking mode. In this case, the replacement trajectory ETR comes without a direction change section RWTA (see Fig. 3 or 5 ). Instead, the equivalent trajectory ETR exhibits a negative curvature in an initial section NTA1 and an end section NTA2 compared to the curvature of the strongly curved subsection TA. The initial section NTA1 extends from the equivalent starting point EA to an inflection point W1, where the curvature becomes positive again. The end section NTA2 extends from an inflection point W2, where the curvature becomes negative, to the equivalent end point EE.

[0063] It can be seen that the replacement trajectory ETR does not exceed the trained trajectory TR in a direction parallel to the initial travel direction DIR of the vehicle 100. Furthermore, a maximum permissible offset DT (see Fig. 2 or 3 ) to the trained trajectory TR is not exceeded (not shown).

[0064] Fig. 5 shows a schematic representation of a fourth example of a replacement trajectory ETR. In this example, two obstacles O, which can be fixed or mobile obstacles, are located next to the trained trajectory TR. The trained trajectory TR has a strongly curved section TA, which describes a 180° curve, followed by a section that describes a 90° curve, the curvature of which does not exceed the maximum permissible curvature. Replacement trajectories ETR, as described above with reference to the Fig. 3 and 4described are not possible because this could lead to a collision with one of the obstacles O. This is indicated by the trajectories X1, X2. Therefore, the replacement trajectory ETR is determined as follows. The replacement starting point EA here corresponds to the starting point A of the strongly curved section TA. A first replacement section ETA1 runs outside the strongly curved section TA up to a first reversal point U1. This is followed by a direction change section RWTA up to a second reversal point U2. In a second replacement section ETA2, the trained trajectory TR is intersected, so that the second replacement section ETA2 runs partly outside and partly inside the strongly curved section TA.

[0065] This example shows that the replacement trajectory ETR can be used with the described concepts, especially with sections of negative curvature NTA1, NTA2 (see Fig. 4 ) as well as with change of direction sections RWTA, can be generated or constructed as required. This allows for a very high degree of flexibility in finding a suitable replacement trajectory ETR. Therefore, the replacement trajectory ETR can be successfully determined even in changing situations, such as new and / or mobile obstacles O and the like. In particular, it is ensured that a maximum permissible offset DT (see Fig. 2 or 3 ) to the trained trajectory TR is not exceeded (not shown).

[0066] Fig. 6 schematically shows a trajectory TR with circles of curvature KR1, KR2. The circle of curvature KR has a radius of curvature R1, and the circle of curvature KR2 has a radius of curvature R2. The circles of curvature KR1, KR2 can also be referred to as osculating circles. The curvature of the trajectory TR is negative in the section described by the circle of curvature KR1 compared to the curvature of the trajectory TR in the section described by the circle of curvature KR2.

[0067] A replacement trajectory ETR (see Fig. 2 - 5 ) is considered to be within the strongly curved section TA (see Fig. 2 - 5 ) if it intersects the corresponding circle of curvature KR1, KR2 of the trained trajectory TR, i.e. runs within the circle of curvature KR1, KR2.

[0068] Fig. 7 shows a schematic block diagram of an embodiment of a method for operating a vehicle 100 (see Fig. 1 - 5 ) by means of a parking assistance system 110 (see Fig. 1 or 8 ). The parking assistance system 110 is in a follow-up mode for driving the vehicle 100 along a trained trajectory TR (see Fig. 2 - 6 ) is set up. The vehicle 100 preferably comprises a number of environmental sensors 120, 130 (see Fig. 1 ) to capture the environment 200 (see Fig. 1 ) of the vehicle 100 and for outputting a corresponding sensor signal to the parking assistance system 110. The parking assistance system 100 is then preferably configured to carry out an automatic parking or following maneuver depending on the received sensor signals.

[0069] The method comprises, in a first step S1, obtaining the trained trajectory TR, which is recorded in a training mode in which a manually driven trajectory is recorded as the trained trajectory TR. In the training mode, which is activated by a user of the vehicle 100, for example, by a corresponding input, the parking assistance system 110 continuously records, for example, a position of the vehicle 100 in the environment 200, an arrangement of objects or obstacles O (see Fig. 6 ) relative to the vehicle 100 as well as odometry data of the vehicle 100. The odometry data include, in particular, a current steering angle, an accelerator pedal position, and a brake pedal position. The trained trajectory TR includes at least one strongly curved section TA (see Fig. 2 - 5 ), whose radius of curvature R1, R2 (see Fig. 6 ) is smaller than a radius that can be achieved with the vehicle 100 with a largest permissible steering angle specified in the follow-up mode.

[0070] In a second step S2, a starting point A (see Fig. 2 - 5 ) and an endpoint E (see Fig. 2 - 5 ) of the highly curved segment TA in the trained trajectory TR is determined. This can occur during sensory acquisition or acquisition of the trajectory, or only when the trained trajectory TR is selected or activated for tracking.

[0071] In a third step S3, a replacement trajectory ETR (see Fig. 2 - 5 ) which has a substitute starting point EA (see Fig. 2 - 5 ) with a substitute endpoint EE (see Fig. 2 -5 ), where the substitute starting point EA and the substitute end point EP each lie on the trained trajectory TR, and the strongly curved section TA lies between the substitute starting point EA and the substitute end point EE. Furthermore, a maximum curvature of the substitute trajectory ETR has a radius of curvature R1, R2, which can be achieved with the largest permissible steering angle specified in the tracking mode or a smaller steering angle. The substitute trajectory ETR determined in this way can therefore be followed without problems in the tracking mode.

[0072] Fig. 8 shows a schematic block diagram of an example of a parking assistance system 110. The parking assistance system 110 is, for example, integrated into a vehicle 100 (see Fig. 1 ), such as a passenger car or a truck. The parking assistance system 110 comprises a computing unit 111 for obtaining a trained trajectory TR (see Fig. 2 - 6 ), which is a trajectory driven manually during a training mode, wherein the trained trajectory TR contains at least one strongly curved section TA (see Fig. 2 -5 ) whose radius of curvature R1, R2 (see Fig. 6 ) is smaller than can be achieved with the vehicle 100 with a maximum permissible steering angle specified in the follow-up mode. Furthermore, a determination unit 112 for determining a starting point A (see Fig. 2 - 5 ) and an endpoint E (see Fig. 2 - 5 ) of the strongly curved section TA in the trained trajectory TR and a determination unit 113 for determining a replacement trajectory ETR (see Fig. 2 - 5 ). The determined substitute trajectory ETR connects a substitute starting point EA (see Fig. 2 - 5 ) with a substitute endpoint EE (see Fig. 2 - 5 ), each of which lies on the trained trajectory TR, with the strongly curved section TA lying between the substitute starting point EA and the substitute end point EE. A maximum curvature of the substitute trajectory ETR has a radius of curvature R1, R2 (see Fig. 6 ) that can be achieved with the largest permissible steering angle specified in the follow-up mode.

[0073] The respective unit, for example, the computing unit 111, the determination unit 112, and / or the detection unit 113, as well as the parking assistance system 110, can be implemented in hardware and / or software. In a hardware implementation, the unit can be configured, for example, as a computer or a microprocessor. In a software implementation, the unit can be configured as a computer program product, as a function, as a routine, as part of a program code, or as an executable object.

[0074] The parking assistance system 110 further preferably has interfaces (not shown) to environmental sensors 120, 130, which are arranged on the vehicle 100 and / or in the environment 200. The parking assistance system 110 receives sensor signals that enable the parking assistance system 110 to accurately estimate the position and / or attitude of the vehicle 100, and based on which the parking assistance system 110 controls the vehicle 100. The determination unit 113 is preferably also configured to take the received environmental sensor signals into account when determining the substitute trajectory ETR.

[0075] Although the present invention has been described using exemplary embodiments, it can be modified in many ways. LIST OF REFERENCE SYMBOLS

[0076] 100Vehicle 110Parking assistance system 111Calculation unit 112Determination unit 113Determination unit 120Optical sensor 130Ultrasonic sensor AStarting point APStarting position DIRDirection of travel DDTDistance EEnd point EASubstitute starting point EESubstitute end point EPEnd position ETA1Substitute sub-section ETA2Substitute sub-section ETRSubstitute trajectory KR1Curvature circle KR2Curvature circle NTA1Starting section NTA2End section OHistor PPParking position R1Radius of curvature R2Radius of curvature RWTADirection of travel change sub-section S1Process step S2Process step S3Process step TAStrongly curved sub-section TRTrained trajectory U1Turning point U2Turning point W1Turning point W2Turning point X1Collision trajectory X2Collision trajectory

Claims

1. Method for operating a vehicle (100) by means of a parking assistance system (110) which is configured, in a following mode, to drive the vehicle (100) along a trained trajectory (TR) or to initiate driving, wherein the method comprises: a) receiving (S1) the trained trajectory (TR) which in a training mode in which a manually navigated trajectory is recorded as the trained trajectory (TR), wherein the trained trajectory (TR) comprises at least one strongly curved section (TA), the radius of curvature (R1, R2) of which is smaller than a radius that can be achieved with the vehicle (100) with a maximum permissible steering lock predefined in the following mode, b) determining (S2) a starting point (A) and an end point (E) of the strongly curved section (TA) in the trained trajectory (TR), c) ascertaining (S3) a substitute trajectory (ETR) which connects a substitute starting point (EA) to a substitute end point (EE), wherein the substitute starting point (EA) and the substitute end point (EE) are each on the trained trajectory (TR) and the strongly curved section (TA) is between the substitute starting point (EA) and the substitute end point (EE), and a maximum curvature of the substitute trajectory (ETR) has a radius of curvature (R1, R2) that can be achieved with the maximum permissible steering lock predefined in the following mode or with a smaller steering lock.

2. Method according to Claim 1, characterized in that the substitute trajectory (ETR) runs at least in sections within the strongly curved section (TA), and / or in that the substitute trajectory (ETR) has a section whose curvature is negative in comparison with the curvature of the strongly curved section (TA), and / or in that the substitute trajectory (ETR) comprises a number of travel direction change sections (RWTA).

3. Method according to Claim 1 or 2, characterized in that the substitute starting point (EA) is before the starting point (A) of the strongly curved section (TA) on the trained trajectory (TR) in relation to a travel direction (DIR) of the vehicle (100), and / or in that the substitute end point (EA) is after the end point (E) of the strongly curved section (TA) on the trained trajectory (TR) in relation to the travel direction (DIR) of the vehicle (100).

4. Method according to one of the preceding claims, characterized in that the substitute trajectory (ETR) is shorter than the strongly curved section (TA).

5. Method according to one of the preceding claims, characterized in that a maximum offset (DT) between the trained trajectory (TR) and the substitute trajectory (ETR) is smaller than a predefined limit value.

6. Method according to one of the preceding claims, characterized in that the substitute trajectory (ETR) comprises at least one travel direction change section (RWTA) whose curvature is smaller than the curvature of the substitute trajectory (ETR) at a section starting point (U1) of the travel direction change section (RWTA), preferably zero, preferably negative.

7. Method according to one of the preceding claims, characterized in that an opening angle between a first tangent to the substitute trajectory (ETR) at the section starting point (U1) of the respective travel direction change section (RWTA) of the number and a second tangent to the trained trajectory (TR) at a point corresponding to the section starting point (U1) is greater than or equal to a predefined limit value.

8. Method according to one of the preceding claims, characterized in that the substitute trajectory (ETR) contains a travel direction change section (RWTA) in each case after a predefined distance along the substitute trajectory (ETR).

9. Method according to one of the preceding claims, characterized in that the substitute trajectory (ETR) has a negative curvature with respect to the strongly curved section (TA) in a starting portion (NTA1), which directly follows the substitute starting point (EA), and / or in an end portion (NTA2) which directly follows the substitute end point (EE).

10. Method according to one of the preceding claims, characterized in that the substitute starting point (EA), the substitute end point (EE) and the substitute trajectory (ETR) are ascertained by means of an iterative optimization method, wherein environmental sensor data which are provided by environmental sensors (120, 130) of the vehicle (100) and / or by external environmental sensors are taken into account in the ascertainment.

11. Method according to one of the preceding claims, characterized in that the vehicle (100) is localized at a localization position on the trained trajectory (TR) which is in front of the starting point (A) of the strongly curved section (TA).

12. Method according to one of the preceding claims, characterized in that the maximum permissible steering angle predefined in the following mode is different for left and right.

13. Computer program product comprising instructions that, when the program is executed by a computer, cause said computer to perform the method according to one of Claims 1 - 12.

14. Parking assistance system (110) for a vehicle (100) which is configured, in a following mode, to drive the vehicle (100) along a trained trajectory (TR) or to initiate driving, the parking assistance system (110) comprising: a computing unit (111) for receiving the trained trajectory (TR) which is a trajectory navigated manually during a training mode, wherein the trained trajectory (TR) comprises at least one strongly curved section (TA), the radius of curvature (R1, R2) of which is smaller than can be achieved with the vehicle (100) with a maximum permissible steering lock predefined in the following mode, a determination unit (112) for determining a starting point (A) and an end point (E) of the strongly curved section (TA) in the trained trajectory (TR), and an ascertainment unit (113) for ascertaining a substitute trajectory (ETR) which connects a substitute starting point (EA) to a substitute end point (EE), wherein the substitute starting point (EA) and the substitute end point (EE) are each on the trained trajectory (TR) and the strongly curved section (TA) is between the substitute starting point (EA) and the substitute end point (EE), and a maximum curvature of the substitute trajectory (ETR) has a radius of curvature (R1, R2) that can be achieved with the maximum permissible steering lock predefined in the following mode or with a smaller steering lock.

15. Vehicle (100) having a parking assistance system (110) according to Claim 14.