Driver assistance system for performing a turning maneuver
The driver assistance system addresses the challenge of automating turning maneuvers by using a sensor system and trajectory planning to identify clear surfaces and execute turning maneuvers automatically, enhancing safety and efficiency.
Patent Information
- Application Number
- DE102023211676
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2023-11-23
- Publication Date
- 2025-06-05
- Estimated Expiration
- 2043-11-23
AI Technical Summary
Existing driver assistance systems for turning maneuvers require a driver and are not capable of automatically executing a turning maneuver, especially in confined spaces like building rooms or parking garages, where visibility and stability issues are prevalent.
A driver assistance system equipped with a sensor system for creating a historical environment map and a trajectory planning system that identifies clear surfaces and plans a turning maneuver in multiple trajectory steps, allowing for automatic execution and continuous feasibility checks.
Enables the automatic execution of turning maneuvers without a driver, reducing the risk of accidents and saving time, while efficiently using unobstructed surfaces and maintaining vehicle stability.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
The invention relates to a driver assistance system for carrying out a turning maneuver of an ego vehicle, the driver assistance system comprising a sensor system for continuously producing a historical environment map, wherein the environment map comprises a past environmental section of the environment in the direction of movement of the ego vehicle, wherein the past section is designed as a section past in time or with respect to the distance travelled.Turning vehicles is often problematic, particularly in rooms for buildings, since typically little space is available, for example due to high traffic volume, parking vehicles or even in parking garages. In most cases, vehicles are moved backward at low speed. The low speed is due to the rearward direction of travel, viewed from the position of the vehicle driver, whereby the visibility conditions are limited. Moreover, the stability of the vehicle presents a problem since the front wheels and thus the rear wheels in the rear direction of travel are steered.The exact process of turning, including preparing and governing dynamics, is dependent on the environment and traffic and can be learned only over a lot of experience. Inextension or impairment can result in over challenge and dangerous situations.Driver assistance systems already exist in the prior art for turning maneuvers, which however require an executing driver in most cases.DE 10 2018 001 248 A1 discloses a driver assistance system for carrying out a turning maneuver of a vehicle having a number of sensors for detecting a vehicle environment, an evaluation unit for evaluating information detected by means of the number of sensors, for determining an available turning space for carrying out the turning maneuver and for determining a number of turning trains required for carrying out the turning maneuver, wherein the evaluation unit, after a steering wheel has been turned in and a predetermined steering angle has been exceeded, determines whether the turning maneuver can be carried out in a turning train, the evaluation unit is coupled to an output device, wherein the output device is configured to output optical and / or acoustic driver information relating to the number of turning trains and then, if more than one turning train is required for carrying out the turning maneuver, to determine whether the turning maneuver is carried out in a turning train, outputting a visual and / or acoustic driver warning.Further driver assistance systems for carrying out a turning maneuver of a vehicle are known from DE 11 2017 003 982 T5 and DE 10 2011 118 551 A1. Methods for monitoring the environment of a vehicle are known from DE 10 2009 047 264 A1 and EP 3 207 538 B1.The object of the invention is to specify a driver assistance system, which is improved compared to the prior art, for automatically carrying out a turning maneuver of a vehicle.The object is achieved by a driver assistance system having the features of claim 1 and an ego vehicle having the features of claim 12.Advantageous embodiments are evident from the dependent patent claims, the description and from the figures.The object is achieved by a driver assistance system for carrying out a turning maneuver of an ego vehicle, the driver assistance system comprising a sensor system for continuously creating a historical environment map, wherein the environment map comprises a past environmental section of the environment in the direction of movement of the ego vehicle, wherein the past section is designed as a section past in time or with respect to the distance travelled, wherein a trajectory planning system is provided which is designed to identify a clear surface on the basis of the historical environment map, and wherein the trajectory planning system is designed to plan a turning maneuver for turning the ego vehicle on the basis of the identified clear surface directly and / or indirectly, wherein the turning maneuver comprises at least a plurality of trajectory steps, wherein the trajectory planning system is designed to initiate an execution of the turning maneuver when planning is successful, Wherein, when the execution is initiated, the sensor system is designed to generate an updated environment map and wherein the trajectory planning system is designed to carry out a feasibility check of the trajectory steps on the basis of the current environment map for success of the turning maneuver, wherein the trajectory planning system is designed to generate the control data with respect to the individual trajectory steps if success and to execute them automatically in succession, and wherein the sensor system is designed to generate a permanent update of the environment map when the trajectory steps are executed until the turning maneuver is complete, wherein the trajectory planning system is designed to carry out a permanent feasibility check of the trajectory steps still to be executed on the basis of the permanent updated environment map until the turning maneuver is complete, or wherein, in the case of a negative feasibility check, the trajectory planning system is designed to perform a continuous feasibility check of the trajectory steps still to be executed on the basis of the permanent updated environment map until the turning maneuver is complete, To do a break of the turning maneuver.A negative feasibility check describes situations such as obstacles or mangiving clearance etc., which make it impossible to carry out the turning maneuver without damage / damage to the ego vehicle / other road users or objects.Termination can also result in a provisional pause.In this case, a turning maneuver describes a maneuver for carrying out a complete turning maneuver of an ego vehicle by approximately 180°, tolerances being intended to be included here. The turning maneuver is started and ended when the ego vehicle is at a standstill.In this case, a sensor system is provided for continuously creating a historical environmental map, wherein the environmental map comprises a past environmental section of the environment in the direction of movement of the ego vehicle, wherein the past section is designed as a section past in time or in relation to the distance travelled. This means that the vehicle is equipped with a suitable sensor system, e.g. camera, radar, lidar, so that, for example, intelligent or other software on a control device is able to record the environment as historical environmental data of the vehicle when travelling in the direction of the own movement and to store it in an environmental map or as an environmental map. Thus, for example, the last 50 meters in the direction of travel can be recorded and stored, for example, in a ring memory, the oldest environmental data always being continuously deleted during a new measurement. Likewise, for example, the last 2 minutes of the journey of the ego vehicle in the direction of movement can be recorded as the environment.In this case, the environment map is preferably created in a local map system / coordinate reference system of the ego vehicle. The environmental data are stored and are thus available as historical data in the form of an environmental map already at the beginning of the turning in order to be able to calculate the complete turning maneuver in advance.Furthermore, according to the invention, a trajectory planning system is provided, for example, as intelligent software, which is designed to identify a clear surface on the basis of the historical map of the surroundings; this can be accomplished, for example, on the basis of the fixed and movable obstacles identified during travel. The clearance surface can be generated on the basis of these obstacles.The trajectory planning system is then designed to plan a turning maneuver for turning the ego vehicle directly and / or indirectly on the basis of the recognized free surface, wherein the turning maneuver comprises at least a plurality of trajectory steps. In this case, direct planning can be carried out, for example, on the basis of an AI module / system present in the trajectory planning system. Furthermore, as indirect planning, the recognized free area and / or the historical environment map can be transmitted to an external server in order to plan the turning maneuver. Both may also be possible. As a result, for example, in the case of a connection to an external server, rapid indirect planning can be made possible; for example, in the parking garage without a connection, direct planning of the turning maneuver can be made possible.The sensor system then creates an updated environment map, wherein the trajectory planning system is designed to carry out a feasibility check of the planned trajectory steps on the basis of the current environment map for success of the turning maneuver.Subsequently, the trajectory planning system according to the invention can generate the control data with respect to the individual trajectory steps and execute them independently in succession when planning / feasibility checking is successful. In parallel, the sensor system is designed to generate a permanent update of the environment map until the turning maneuver has been completed. For safety reasons and for the case of turning in a traffic-calm zone, the maximum speed during the turning maneuver may be limited, for example to 7 km / h.In parallel, a permanent feasibility check of the trajectory steps still to be carried out is accomplished on the basis of the permanent updated environment map. This can also be done, for example, by including the current environment map in the generation of the individual trajectory steps.A feasibility check is carried out until the turning maneuver is completed or the turning maneuver is interrupted in the case of a negative feasibility check.The driver can save both the outlay for planning and the time outlay for turning the ego vehicle and can also reduce the risk of an accident by overloading by the driver assistance system according to the invention. The driver assistance system measures and stores obstacle data and efficiently uses the unobstructed surface to turn and restore the ego vehicle to the previous position.For safety reasons and in the case of turning in a traffic-calm zone, the maximum speed during the turning maneuver can be limited to 7 km / h or kept very low.In this case, the turning maneuver is started and ended when the ego vehicle is at a standstill.The driver assistance system according to the invention enables an automated turning maneuver to be carried out even without a driver. This results in a reduced risk of an accident and a high time saving.In a further embodiment, an interface is provided, wherein the trajectory planning system is designed to transmit the recognized free surface and / or the map of the surroundings to an external server, for planning a turning maneuver for turning the ego vehicle in a plurality of trajectory steps as indirect planning, wherein the trajectory planning system is designed to receive the indirect planning of the turning maneuver determined by the server using the trajectory steps on the basis of the interface.Subsequently, for example, the generated turning maneuver can be matched to an internally generated turning maneuver and executed only if they match. This means that the environment data and / or the recognized free area for movement planning are sent via the connection to a suitable cloud server, which plans a trajectory planning for turning using a deep learning algorithm, for example. In this case, such a cloud server can use not only the data of the ego vehicle, but also other external data from other road users or data from other infrastructures, and also, for example, an HD map. As a result, a turning maneuver can be planned reliably. Furthermore, in the case of a poor connection, for example, only the turning maneuver planned in the ego vehicle can be used.In a further embodiment, a turn activation element is provided for manually activating the turn or for manually activating the turn in a time-delayed manner. For this purpose, a display can be provided, for example, on which the driver or the user can achieve manual activation. In this case, the manual activation can also comprise a time window and a time duration in which the turning maneuver is to take place. In this case, for example, the user can activate the driver assistance system only when the ego vehicle is at a standstill of the vehicle.Alternatively or optionally, the trajectory planning system is designed to receive a turning signal from an authenticated source and to initiate a turning maneuver immediately or in a time-shifted manner on the basis of the received turning signal. An authenticated source, for example a mobile terminal of the driver or of the user, is thereby one which can be identified by the ego vehicle. In this case, such authentication corresponds substantially to an identification.In a further embodiment, the trajectory planning system is designed to transmit a success signal when the turning maneuver is successfully planned. After a successful planning of the turning maneuver on the basis of the environment map and the clearance area calculated therefrom, the driver assistance system can transmit an activation recommendation to the user. This transmission can be displayed on a display, for example, or else on a mobile terminal of the user. The user then has the possibility of setting a timer for the time-delayed starting of the turning maneuver or of initiating an immediate turning maneuver. In the event of a failure to confirm by the user without a set timer, the driver assistance system can deactivate itself again, for example.In a further embodiment, the trajectory planning system defines a starting position and a target position, wherein the target position lies in the region opposite the starting position. In this case, for example, the target position can be rotated by 180° with respect to the starting position, which corresponds to a complete turning maneuver. Tolerances can also be established or have been established within which a turning maneuver is considered complete.In particular, the trajectory planning system is designed to transmit the target position to an authenticated source. In this case, the transmission can be effected by display on the display or, for example, by display on the display of a mobile terminal of the user as a source. Furthermore, the trajectory planning system is designed for the subsequent reception of an acknowledgment signal. In particular, after the confirmation signal is received by the authenticated source, the turning maneuver can be initiated.According to the invention, the trajectory planning system is designed to replace the target position with the original starting position when the turning maneuver is interrupted and to initiate a return maneuver.In a further embodiment, the trajectory planning system is designed to, when the turning maneuver is discontinued, execute the past trajectory steps which have already been executed and have been partially executed in the opposite direction as a return maneuver in order to reach the previous starting position. This provides a possibility for pulling back in the event of failure, on the one hand to enable repetition and, on the other hand, so that the ego vehicle does not remain on the roadway and obstructs other road users.According to the invention, the trajectory planning system is designed to initially interrupt or pause the turning maneuver for a predefined time duration when an obstacle is detected, and to abort the turning maneuver only after the predefined time duration when the obstacle is continuously detected. This can prevent premature termination. The ego vehicle pauses until, for example, an obstacle that suddenly appears in the driving area, e.g., a pedestrian, is no longer detected and successful execution of the turning maneuver is possible.Furthermore, in a further embodiment, the trajectory planning system can be designed to carry out the past trajectory steps which have already been carried out and have been partially carried out in the opposite direction as a return maneuver in order to reach the previous starting position when the turning maneuver is interrupted. This allows simple return to the starting position.According to the invention, the trajectory planning system is designed to plan a return maneuver for returning the ego vehicle on the basis of the open area on the basis of the original starting position as a new target position, wherein the return maneuver comprises a plurality of trajectory steps, and wherein the trajectory planning system is designed to initiate an execution of the return maneuver when planning is successful, and wherein the sensor system is designed to generate an updated environment map when the execution is initiated, and wherein the trajectory planning system is designed to perform a feasibility check of the trajectory steps on the basis of the current environment map for success of the return maneuver, wherein the trajectory planning system is designed to generate the control data with respect to the individual trajectory steps when success and to execute them successively, and wherein the sensor system is designed to generate a permanent update of the environment map when the trajectory steps are executed, until the return maneuver is completed, wherein the trajectory planning system is designed to perform a permanent feasibility check of the trajectory steps still to be carried out on the basis of the permanent updated environment map until the return maneuver is completed.Such a return maneuver can take into account newly occurring obstacles. Furthermore, the return maneuver can likewise be accomplished first by executing the original trajectory steps in opposite directions, and a new return maneuver can be calculated only if it fails.Furthermore, the trajectory planning system can be designed to assume an original operating state when the return maneuvers have taken place. After reaching the starting position, the driver assistance system thus automatically sets itself back to stand-by. Thus, a new attempt can be started at a later time, wherein the ego vehicle is at a safe location again until then.In a further embodiment, the trajectory planning system is designed to bring about an abort of the return maneuver in the event of a negative feasibility check. Thus, a new attempt can be started at a later time, wherein the ego vehicle is at a safe location again until then.In a further embodiment, the trajectory planning system has a trained AI module which is designed at least for planning a turning maneuver for turning the ego vehicle in a plurality of trajectory steps. In this case, the AI module can be, for example, a neural network which has been trained on the basis of clearance surfaces in order to produce a turning maneuver. Such training can be carried out, for example, on the basis of simulation data and / or real data. However, other deep learning algorithms can also be used. In particular, the driver assistance system can be designed to train the AI module further on the basis of manually executed turning maneuvers. Furthermore, an update of the AI module can likewise take place on the basis of a cloud server, with the result that the AI module experiences a permanent improvement, as a result of which the safety is increased.This allows a number of different scenarios to be taken into account, which can be extended and improved by the learning capability in combination with other systems.In this case, the AI module can also be designed to identify a clear surface on the basis of the environment map.In a further embodiment, a V2X interface is provided for receiving external environment data in the area of the ego vehicle, for enriching the historical environment data recorded by the sensor system and / or the current environment data. These may be, for example, the environmental data from a cloud / edge server / drone / traffic monitor and / or V2V data from surrounding vehicles that are used to create a more accurate environmental map or verify the created one. This makes it possible to produce a safe turning maneuver.In a further embodiment, the trajectory planning system can be designed to plan the turning maneuver at least on the basis of a maximum number and an order of the trajectory steps to be carried out and also the required time. Simple abort criteria can thereby be created. If the maximum number of trajectory steps is exceeded and / or if, for example, there is an excessively long standstill in front of an obstacle, or if the withdrawal is activated by the user, etc., the turning maneuver can be interrupted and the target position and target alignment can be set as new starting values, i.e. the original starting values are used as new target values.The object is furthermore achieved by an ego vehicle having a driver assistance system as described above. The ego vehicle may be an autonomously or semi-autonomously operated vehicle that has corresponding sensor systems and corresponding computer capacities.Further characteristics and advantages of the present invention will become apparent from the following description with reference to the attached figures. The following are shown schematically in the drawing: FIG. 1 : a driver assistance system according to the invention schematically, FIG. 2 : an ego vehicle according to the invention with a sensor system, FIG. 3 : a generation of a first turning maneuver with trajectory steps, FIG. 4 : a generation of a second turning maneuver with trajectory steps, FIG. 5 : a generation of a third turning maneuver with trajectory steps and obstacles.FIG. 1 shows a driver assistance system 1 according to the invention schematically for carrying out a complete turning maneuver of an ego vehicle 2. a turning maneuver describes a turning / rotation of the ego vehicle 2 by approximately 180°, it being possible for tolerances to be included.The driver assistance system 1 has a sensor system 3, e.g. camera, radar sensors, lidar sensors, so that an evaluation module 7, for example intelligent software, is able to measure the environment of the ego vehicle 2 as environmental data.This means that the environment of the ego vehicle 2 is measured, i.e. the driver assistance system 1 is placed in a recording mode A. On the basis of the measured environment, an environment map can now be generated by the evaluation module 7 as a historical environment map. In this case, the sensor system 3 continuously generates a historical environmental map, wherein the environmental map comprises a past environmental section of the environment in the direction of movement of the ego vehicle 2, wherein the past section is designed as a section past in time or in relation to the distance travelled. Thus, for example, the last 50 meters of the journey of the ego vehicle 2 can be recorded in each case and stored, for example, in a ring memory, the oldest data always being deleted or overwritten during continuous re-measurement. Likewise, for example, the last 2 minutes can be recorded. In this case, the environment map is preferably created in a local map system / coordinate reference system. The environmental data are stored and are thus available as historical data in the form of an environmental map already at the beginning of the turning maneuver in order to be able to calculate the complete turning maneuver in advance.FIG. 2 shows such a sensor system 3 in an ego vehicle 2. the ego vehicle 2 or the driver assistance system 1 can have an interface 4 for receiving external environment data in the area of the ego vehicle 2, for enriching the historical environment data recorded by the sensor system 3. These can be, for example, the environmental data from a cloud or edge server / drone 6 / traffic monitoring system 9 and / or V2V / V2X data, from vehicles located in the environment, which are used to create a more accurate environmental map or to verify the created data.As soon as the ego vehicle 2 is in the idle state, i.e. in the idle state, the user can activate the driver assistance system 1. In this case, for example, a turn activation element, for example a display in the ego vehicle 2, can be provided for manually activating the turn or for manually activating the turn with a time delay by the user. In this case, the manual activation can also comprise a time window in which the turning maneuver is to take place or a time (timer) at which the turning maneuver is to take place. Alternatively, for example, the user can transmit an activation signal via a mobile authenticated, i.e. identified, terminal 10.Furthermore, the driver assistance system 1 has a trajectory planning system 8, which is placed in a provision mode B after activation and is designed to detect open spaces on the basis of the historical environment map. Furthermore, in the provision mode B, the turning maneuver for turning the ego vehicle 2 is planned on the basis of the recognized clearance area, wherein the planning of the turning maneuver comprises at least a plurality of trajectory steps. For this purpose, the trajectory planning system 8 has an existing AI module, for example a deep learning algorithm. In this case, the turning maneuver can plan the plurality of trajectory steps, for example using the deep learning algorithm. However, a different AI module or other algorithm can also be used.Alternatively or additionally, the trajectory planning system 8 can transmit the environment data and / or environment map and the recognized open spaces to, for example, an external cloud or edge server 5 for planning the turning maneuver. The planned turning maneuver can then be received by the trajectory planning system 8 for implementation.FIG. 3 shows a generation of a turning maneuver with the trajectory steps T 1, T 2, T 3 from a starting position S to a target position Z in a small space.FIG. 4 shows a generation of a complicated turning maneuver with parking bays P as a clearance area with the trajectory steps T1,T2,T3,T4,T5 from a starting position S to a target position Z.A turning maneuver consists of a number n of trajectory steps T 1,..., Tnand also a time duration.Before the initiation of the turning maneuver, the starting position S and the target position Z are defined with the respective alignment.If the trajectory planning system 8 can provide successful planning, the driver assistance system 1 notifies the user of an activation recommendation.The user has the possibility of setting a timer for the time-delayed starting of the turning maneuver or, for example, starting an immediate turning maneuver by confirming the turning maneuver. In the event of a failure to confirm by the user without a set timer, the driver assistance system 1 can deactivate itself again.For confirmation by the user on site or by the identified terminal 10 or by timer activation, the target position Z can be transmitted in advance to the user who confirms it, for example. The turning maneuver can only be started from standstill.For each trajectory step T 1,..., Tn, the control data are determined in a maneuver mode M and transmitted to those control units 11 which are necessary for carrying out the movement planning such as steering, acceleration and braking, collision avoidance and actuator control.The control data are then scanned by the control units 11. In parallel, the environment is newly and continuously measured by the sensor system 3 to generate environment data.This means that the sensor system 3 generates a permanent update of the environment map during and before execution, i.e. also shortly before the start to the end of the turning maneuver, by the trajectory steps until the turning maneuver has been completed.On the basis of the permanently updated environment map, the trajectory planning system 8 carries out a permanent feasibility check with respect to the trajectory steps to be carried out. If this is positive, the turning maneuver can be carried out. During the execution of the trajectory steps, which are carried out at a low speed, for example, at a maximum of 7 km / h here, sensor system 3 thus detects the environment around ego vehicle 2 and continuously updates the environment map. In parallel, obstacles are sought in the driving area on the basis of the updated environment map.The user can obtain the current data about the progress of the turning maneuver as well as the completion of the turning maneuver. After completion of the completion, hereby identified as E, of the turning maneuver, the driver assistance system 1 can be put back into readiness again.When an obstacle is detected, the driver assistance system 1 can pause the turning maneuver in order to avoid a collision. Here, the detection of an obstacle corresponds to a negative feasibility check. After the pause, the turning maneuver can be continued.If the feasibility check continues to be negative, and if the turning maneuver after the start cannot be ended for various reasons despite a previous activation recommendation, the trajectory planning system 8 carries out an abort of the turning maneuver. Termination criteria can be, for example: invalid trajectory, exceeding the maximum number of trajectory steps and / or an excessively long standstill in front of an obstacle, or activation of the withdrawal by the user etc.The previously carried out turning maneuver is then reversed in a retraction mode R.For this purpose, the trajectory planning system 8 can set the target position and target alignment as new starting values, i.e. the original starting values are used as new target values.Furthermore, when the turning maneuver is discontinued, the past trajectory steps which have already been carried out and have been partially carried out can be carried out in the opposite direction as return maneuvers for reaching the previous starting position. This allows simple return to the starting position, wherein the sensor system 3 likewise generates a permanently updated environment map here as well. On the basis of the permanently updated environment map, the trajectory planning system 8 carries out a permanent feasibility check for obstacle detection with respect to the trajectory steps to be carried out in the retraction mode R.Alternatively, the trajectory planning system 8 can generate the returning maneuver using a plurality of trajectory steps for returning the ego vehicle 2 using the clear surface on the basis of the original starting position as a new target position.Subsequently, when planning is successful and the user confirms it, execution of the return maneuver can be initiated.On the basis of the permanently updated environment map, the trajectory planning system 8 carries out a permanent feasibility check with respect to the trajectory steps to be carried out. During the execution of the trajectory steps, which are carried out at a low speed, for example, at a maximum of 7 km / h here, the sensor system 3 detects the environment around the ego vehicle 2 and continuously updates the environment map. In parallel, obstacles are sought in the driving area on the basis of the updated environment map. After reaching the original starting position, the driver assistance system 1 automatically puts itself back into readiness. Thus, a new attempt can be started at a later time, wherein the ego vehicle 2 is again in a safe place until then.FIG. 5 shows a generation of a complicated turning maneuver with obstacles H by the driver assistance system 1 according to the invention and parking bays P as a clear surface.This has the trajectory steps T 1, T 2, T 3 for turning from a starting position S to a target position Z. The driver assistance system 1 according to the invention makes it possible for the user to save both the outlay for planning and the time outlay for turning the ego vehicle 2. This allows the risk of an accident to be reduced by an over challenge. The turning maneuver can be carried out at any desired time with or without the presence of the user and repeated if it fails.The driver assistance system 1 according to the invention also offers the possibility of carrying out a return maneuver in the retraction mode R in the event of failure. This on the one hand allows repetition. On the other hand, the ego vehicle 2 therefore does not stay on the roadway and obstructs other road users.The turning maneuver can likewise pause as long as an obstacle which suddenly appears in the driving range, e.g. a pedestrian, is detected and corrected or until a return maneuver is initiated in the retraction mode R. For safety reasons and for the case of turning in a traffic-calm zone, the maximum speed during the turning maneuver can be limited, for example to 7 km / h.The driver assistance system 1 according to the invention is furthermore designed to improve the itself generated environmental data for generating an environmental map on the basis of the environmental data from a cloud or edge server / drone 6 / traffic monitoring 9 and / or V2V / V2X data.An AI module can also be used which is continuously improved by posttraining or updates. Driver assistance system 1 is therefore trainable, as a result of which efficiency and reliability are constantly increased.Furthermore, the driver assistance system 1 measures and stores obstacle data and efficiently uses the obstacle-free area to turn and turn the ego vehicle 2 in a turned position.The user has the options of cancelling the ongoing turning maneuver in the ego vehicle 2 or by terminal 10 at any time or instructing the withdrawal to the starting position and starting alignment.The turning maneuver can be started at a specific time by the user in ego vehicle 2, via a terminal device 10 or via a programmable timer.List of reference numbers:1 Driver assistance system 2 ego vehicle 3 sensor system 4 interface 5 cloud 6 drone 7 evaluation module 8 trajectory planning system 9 traffic monitoring device 10 terminal 11 control device P parking bays R withdrawal mode H obstacle B provision mode A recording mode M maneuver mode T 1,..., Tn trajectory steps
Claims
Driver assistance system (1) for carrying out a turning maneuver of an ego vehicle (2), the driver assistance system (1) comprising a sensor system (3) for continuously producing a historical environment map, wherein the environment map comprises a past environmental section of the environment in the direction of movement of the ego vehicle (2), wherein the past section is designed as a section past in time or in relation to the distance travelled, wherein a trajectory planning system (8) is provided which is designed for detecting a clear surface on the basis of the historical environment map, and wherein the trajectory planning system (8) is designed to plan a turning maneuver for turning the ego vehicle (2) on the basis of the detected clear surface directly and / or indirectly, wherein the turning maneuver comprises at least a plurality of trajectory steps, wherein the trajectory planning system (8) is designed to plan a turning maneuver for turning the ego vehicle (2) on the basis of the detected clear surface directly and / or indirectly, In the case of successful planning, an execution of the turning maneuver is initiated, and wherein, in the case of initiation of the execution, the sensor system (3) is designed to generate an updated environment map, and wherein the trajectory planning system (8) is designed to carry out a feasibility check of the trajectory steps on the basis of the current environment map for success of the turning maneuver, wherein the trajectory planning system (8) is designed to generate the control data relating to the individual trajectory steps and to execute them successively automatically, in the case of success, and wherein the sensor system (3) is designed to generate a permanent update of the environment map, in the case of execution of the trajectory steps, until the turning maneuver has been completed, wherein the trajectory planning system (8) is designed to carry out a permanent feasibility check of the trajectory steps still to be executed on the basis of the permanent updated environment map until the turning maneuver has been completed, or wherein, in the case of a negative feasibility check, the trajectory planning system (8) is designed to bring about an abort of the turning maneuver, characterized in that the trajectory planning system (8) is designed to replace the target position by the original starting position and to initiate a return maneuver when the turning maneuver is aborted, and in that the trajectory planning system (8) is designed to initially interrupt or pause the turning maneuver for a predefined period of time when an obstacle is detected, and to abort the turning maneuver only after the predefined period of time when the obstacle is detected continuously, and in that the trajectory planning system (8) is designed to plan a return maneuver for returning the ego vehicle (2) as a new target position on the basis of the flank on the basis of the original starting position, wherein the return maneuver comprises a plurality of trajectory steps, and wherein the trajectory planning system (8) is designed to:, In the case of successful planning, an execution of the return maneuver is initiated, and wherein, in the case of initiation of the execution, the sensor system (3) is designed to generate an updated environment map, and wherein the trajectory planning system (8) is designed to carry out a feasibility check of the trajectory steps on the basis of the current environment map for success of the return maneuver, wherein the trajectory planning system (8) is designed to generate the control data relating to the individual trajectory steps and to execute them successively, in the case of success, and wherein the sensor system (3) is designed to generate a permanent update of the environment map, in the case of execution of the trajectory steps, until the return maneuver is completed, wherein the trajectory planning system (8) is designed to carry out a permanent feasibility check of the trajectory steps still to be executed on the basis of the permanent updated environment map until the return maneuver is completedDriver assistance system (1) according to Claim 1, characterized in that an interface (4) is provided, wherein the trajectory planning system (8) is designed to transmit the recognized free surface and / or the map of the surroundings to an external server, for planning a turning maneuver for turning the ego vehicle (2) in a plurality of trajectory steps as indirect planning, wherein the trajectory planning system (8) is designed to receive the indirect planning of the turning maneuver determined by the server using the trajectory steps on the basis of the interface (4).Driver assistance system (1) according to Claim 1 or 2, characterized in that a turning activation element is provided for manually activating the turning or for manually activating the turning in a time-delayed manner and / or wherein the trajectory planning system (8) is designed to receive a turning signal from an authenticated source and to initiate a turning maneuver immediately or in a time-delayed manner on the basis of the received turning signal.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) is designed to transmit a success signal when the turning maneuver is successfully planned.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) specifies a start position and a target position, wherein the target position lies in the region opposite the start position.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) is designed, when the turning maneuver is interrupted, to carry out the past trajectory steps which have already been carried out and have been partially carried out in the opposite direction as a return maneuver in order to reach the previous starting position.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) is designed to assume an original operating state when the return maneuver has taken place.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) is designed to bring about termination of the return maneuver in the event of a negative feasibility check.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) has a trained AI module which has at least for planning a turning maneuver for turning the ego vehicle (2) in a plurality of trajectory steps.Driver assistance system (1) according to one of the preceding claims, characterized in that a V2X interface is provided for receiving external environment data in the region of the ego vehicle (2), for enriching the historical environment data recorded by the sensor system (3) and / or the current environment data.Driver assistance system (1) according to one of the preceding claims, characterized in that the trajectory planning system (8) is designed to plan the turning maneuver at least on the basis of a maximum number and an order of the trajectory steps to be carried out and also the required time.Ego vehicle (2) having a driver assistance system (1) according to one of the preceding claims.
Citation Information
Patent Citations
Method for support of driver of vehicle, involves determining environment of vehicle and determining prospective driving direction at branching, crossing or curve by evaluation of driven trajectory
DE102009047264A1
Method for supporting autonomous or semi-autonomous driving maneuver e.g. off-parking maneuver, of motor car, involves using assistance system when recognizing presence of termination condition in form of user action
DE102011118551A1
driver assistance system for performing a turning maneuver of a vehicle
DE102018001248A1
Systems and procedures for controlling vehicle maneuvers
DE112017003982T5
Communication system for a vehicle and method for communicating
EP3207538B1