Method and parking-assistance device for dynamically planning a parking trajectory of a vehicle during an at least partly automatically performed parking process

Abstract

The invention relates to a method for dynamically planning a parking trajectory of a vehicle (1) during an at least partly automatically performed parking process, wherein the method comprises the following steps: detecting the surroundings of the vehicle by means of at least one sensor device; determining and fixing a target position for the vehicle based on the surroundings data detected by the at least one sensor device; planning a parking trajectory based on the fixed target position; recalculating the target position (Z2) during the parking process based on surroundings data updated during the parking process; comparing the difference between the fixed target position and a formed average value from at least one subset of the target positions recalculated during the parking process; wherein the fixed target position is updated if the difference between the fixed target position and the formed average value exceeds a specified limit value; dynamically replanning the parking trajectory based on the updated target position.

Description

[0001] Description

[0002] Method and parking assistance device for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking operation

[0003] The invention relates to a method for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking process, wherein the surroundings of the vehicle are detected by at least one sensor device, and a target position for the vehicle is determined and specified based on the surroundings data detected and transmitted by the at least one sensor device. Furthermore, the invention relates to a parking assistance device and a vehicle.

[0004] Parking assistance systems are already known in the state of the art which, based on the current driving scenario, represented by an environment model, can calculate a trajectory on which the vehicle moves at least partially automatically, i.e., for example, without steering intervention by the driver and / or without actuation of the accelerator or brake pedal by the driver.

[0005] During the at least partially automated execution of the parking process, changes to the environmental model provided by a driver assistance system of the vehicle can occur. The change in the environmental model can have various causes, for example, that the environmental detection is more accurate due to the lower speed during the parking process (compared to the detection when driving past the parking space), that new environmental details are detected when entering the parking space that were not detected by the vehicle's sensors when driving past the parking space, a change in the parking situation (e.g., the parking space is larger or smaller than when the parking process was originally planned), and / or that a temporal change in the environmental situation occurs, for example, due to a pedestrian stepping onto the edge of the parking space, etc.As a result, it may happen that the initially determined parking trajectory can no longer be carried out without collisions or that the vehicle takes up an unfavorable or inaccurately aligned target final position.

[0006] In such situations, known parking assistance systems terminate the parking process or at least stop the vehicle. Alternatively, an update of the calculated target position based on the new environmental details is conceivable in order to arrive at a more optimal parking position. However, this can result in constant updating of the target position, e.g. in each measurement cycle of a sensor unit. The change in the target position is associated with a recalculation of an alternative parking trajectory, which often makes maneuvering necessary, i.e. the vehicle is moved along a multi-lane parking trajectory rather than a single-lane one. This type of replanning is perceived as uncomfortable by the vehicle occupants.

[0007] Based on this, it is the object of the invention to provide a method for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking process, which can react to a changed parking situation and has a high level of parking comfort and thus a high level of user acceptance.

[0008] This object is achieved by a method having the features of independent patent claim 1. Preferred embodiments are the subject of the dependent claims. A parking assistance device for planning a parking trajectory of a vehicle is the subject of independent patent claim 9.

[0009] According to a first aspect, the invention relates to a method for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking process, in particular entering or exiting a parking space. The vehicle can be, for example, a passenger car (PC) with four wheels. The vehicle can also be configured as another type of motor vehicle or as a non-motorized vehicle, for example, as a trailer. In particular, the vehicle can also be a combination of a motor vehicle and a trailer. The method comprises the following steps:

[0010] First, the vehicle's surroundings—for example, a parking space, a parking space, or an inductive charging plate—are detected by at least one sensor device of the vehicle. Based on the surrounding data detected by the at least one sensor device, a target position for the vehicle is determined and specified. The parking trajectory is then planned and, in particular, controlled based on the specified target position. In particular, the parking trajectory describes a movement of the vehicle toward the target position. The surrounding data includes, for example, information from a road model, object information in the vehicle's surroundings, and a driving dynamics model of the vehicle.

[0011] During the at least partially automated parking process along the planned, initial, or most recently defined parking trajectory, the target position is recalculated based on environmental data updated during the parking process. The recalculation of the target position serves to determine whether the vehicle should be moved to a new target position that differs from the initial target position. Particularly preferably, the recalculation is performed continuously, i.e., at regular intervals.

[0012] After the target position has been recalculated, a difference between the initially or most recently determined target position is compared with an average calculated from at least a subset of the target positions recalculated during the parking process. Outliers are preferably ignored when calculating the average from the recalculated target positions.

[0013] The initially or last specified target position is updated if the difference between the initially or last specified target position and the calculated average exceeds a specified limit. In other words, the parking situation and / or the location of the newly calculated target positions relative to the initially or last specified target position is assessed and adjusted based on the assessment result. The parking trajectory is dynamically replanned based on the updated target position.

[0014] When recalculating the target positions, it is possible that the automated parking process has not yet begun, for example, that an updated surrounding area was already detected when parking began. However, it is also possible that the target positions are recalculated after part of the current parking trajectory has already been traversed.

[0015] The method according to the invention can advantageously determine whether a new target position is manifested by the set of target positions recalculated during the parking process or whether new target positions arise solely due to measurement tolerances. At the same time, larger deviations can be reliably determined by averaging the newly calculated target positions, allowing a time-efficient response. In the event of minor changes in the environment, however, the target position and parking trajectory are maintained, ensuring robust automated parking behavior. This prevents constant replanning of the parking trajectory, which requires high computational complexity.

[0016] According to a preferred embodiment, the target positions recalculated during the parking process are weighted according to their recency when calculating the mean. In particular, the more recent data are given a higher weighting to enable the target position to be adapted to the current environment.

[0017] Should it be necessary to update the specified target position, it is preferable that the most recently recalculated target position is selected as the updated target position. This is advantageous, for example, if there is a slight deviation from the previous target positions and it can be assumed that the most recently calculated target position indicates a precise value. However, it can also be provided that the calculated mean value is selected as the updated target position. Preferably, the mean value is calculated from at least three newly calculated target positions, which, however, do not necessarily have to be evaluated consecutively. Rather, it is preferable to use the mean value of at least three newly calculated target positions that take into account a parking space that is as completely recorded as possible, for example the entire lateral boundary of the parking space.In this way, the need to update the target position again due to an insufficient or incompletely recorded environment can be reduced.

[0018] Particularly preferably, the specified target position comprises a position and an orientation of the vehicle relative to a spatially fixed coordinate system, wherein the updating of the target position can relate to the position and / or the orientation. Thus, the check as to whether the specified target position needs to be updated includes, for example, determining the difference between a specified lateral target position value and the newly calculated lateral target position values.

[0019] According to one embodiment, the target position is recalculated in each measuring cycle of the at least one sensor device during the parking process.

[0020] According to one embodiment, the parking maneuver is a diagonal parking maneuver or a perpendicular parking maneuver. These parking maneuvers are characterized by the fact that, due to the large parking space depth, the initial detection of the parking situation when driving past the parking space is often insufficient. Upon at least partially entering the parking space, the surrounding model is updated, necessitating replanning the initial parking trajectory.

[0021] According to one embodiment, the speed of the vehicle when driving through the specified parking trajectory is between 0.2 m / s and 3 m / s. Due to the low vehicle speed, the updated parking trajectory can have relatively small curve radii, so that the new target position can be reached even with a relatively short new parking trajectory.

[0022] According to a further aspect, the invention relates to a parking assistance device for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking process. The invention is particularly advantageous for automation levels of SAE Level 2 and higher.

[0023] The parking assistance device comprises a sensor device for detecting the surroundings of the vehicle, as well as an evaluation unit configured to determine and specify a target position for the vehicle based on the surroundings data detected and transmitted by the at least one sensor device. Furthermore, the evaluation unit is configured to plan a parking trajectory based on the specified target position. The parking trajectory is, in particular, a combination of circular arc segments, straight lines, and / or clothoids, by means of which the specified target position is reached.

[0024] The evaluation unit is configured to recalculate target positions during the parking process based on the environmental data updated during the parking process. Furthermore, the evaluation unit is configured to compare the difference between the specified target position and a calculated mean value from at least a subset of the target poses recalculated during the parking process and to output an update of the specified target position if the difference between the specified target position and the calculated mean value exceeds a predetermined limit value. The evaluation unit is configured to dynamically replan the parking trajectory based on the updated target position. In particular, a computing unit is configured to automatically guide the vehicle to be parked along the newly planned parking trajectory.

[0025] According to a further aspect, the invention relates to a motor vehicle with a parking assistance device according to the preceding description. The terms "approximately," "substantially," or "about" within the meaning of the invention mean deviations from the respective exact value of + / - 10%, preferably + / - 5%, and / or deviations in the form of changes that are insignificant for function.

[0026] Further developments, advantages, and possible applications of the invention will become apparent from the following description of exemplary embodiments and from the figures. All described and / or illustrated features, individually or in any combination, are fundamentally part of the invention, regardless of their summary in the claims or their reference back to them. The content of the claims is also incorporated into the description.

[0027] The invention is explained in more detail below with reference to exemplary embodiments and the figures. They show:

[0028] Figure 1 shows an example of a schematic representation of a parking situation with an initially defined target position;

[0029] Figure 2 shows an example of the parking situation with an updated

[0030] Target position based on which the new target position is reached instead of the last set target position of the vehicle;

[0031] Figure 3 shows an exemplary series of measurements of newly calculated lateral

[0032] target position values;

[0033] Figure 4 shows an example of a block diagram showing the process steps of the

[0034] Procedure for dynamically replanning the defined parking trajectory is illustrated.

[0035] The present invention relates to a method for dynamically planning a parking trajectory of a vehicle during an at least partially automated parking process by a parking assistance device. "Dynamic planning" means that, after determining a new target position that takes into account the changed environmental situation of the vehicle, the vehicle is moved further along a parking trajectory adapted to the new target position. The trigger for replanning the parking trajectory can be automatically generated by an evaluation unit of the vehicle, for example, when a sensor device of the vehicle detects that the parking situation has changed and the vehicle must assume a new target position.Such a change in the parking situation occurs, for example, if the parking space that is to be parked in by means of the parking process has increased or decreased in size, an object that was not initially recognized limits the parking space, or the parking space orientation was initially assumed to be different.

[0036] Figure 1 shows, by way of example and schematically, an initially defined target position Z1 for a parking maneuver of a vehicle 1 that is to be parked in a parking space PL. The parking space PL is bounded, for example, by objects such as third-party vehicles, curbs, or boundary lines.

[0037] For example, due to the speed of vehicle 1 when passing the parking space PL and / or the distance of vehicle 1 from the parking space PL, it may happen that the parking trajectory is planned based on an environmental model that does not correspond to the actual prevailing parking situation. For example, an obstacle that prevents vehicle 1 from positioning at the initially defined target position Z1 may not be detected until entering the parking space, necessitating replanning of the initially defined target position Z1 to a new target position Z2 and, consequently, a new parking trajectory.

[0038] The first target position Z1 is generated, in particular, based on an environmental model created before the start of the parking process. After the parking process has started, and in particular after vehicle 1 is moved along the initial parking trajectory by the parking assistance device, the target position is recalculated at regular intervals based on environmental data updated during the parking process.

[0039] Figure 2 shows an updated target position Z2, which differs from the initial target position Z1 and to which the vehicle 1 is to be maneuvered along a newly planned parking trajectory.

[0040] Figure 3 shows, purely by way of example, seven newly calculated lateral target positions Z2 - Z8. An average is calculated from at least a subset of these newly calculated target positions Z2 - Z8. In this exemplary embodiment, measured value 3 is an outlier and is therefore not taken into account when calculating the average. The target positions Z2 - Z8 are recalculated, for example, in each measurement cycle of the at least one sensor device of vehicle 1.

[0041] To determine with minimal computational effort whether an update of the planned parking trajectory is necessary, a difference between the specified target position Z1 and a calculated mean value from at least a subset of the newly calculated target positions Z2 - Z8 is compared. If the difference between the specified target position Z1 and the calculated mean value exceeds a specified limit, the specified target position Z1 is updated. This achieves a dynamic and robust correction of target positions, thus avoiding disadvantageous parking positions.

[0042] Fig. 4 shows a schematic block diagram illustrating the dynamic planning of the parking trajectory of the vehicle 1 during an at least partially automated parking process. First, the surroundings of the vehicle 1 are detected S10 by at least one sensor device. In a next step, the target position Z1 for the vehicle 1 is determined and defined S11 based on the surroundings data detected by the at least one sensor device, with the parking trajectory being planned S12 based on the defined target position. The target position is recalculated during the parking process based on surroundings data updated S13 during the parking process.In a next step, the difference between the specified target position and a calculated mean value from at least a subset of the target positions Z2 - Zx recalculated during the parking process is compared. The specified target position is updated if the difference between the specified target position and the calculated mean value exceeds a specified limit value S14. The parking trajectory is dynamically replanned based on the updated target position S15.

[0043] The method described above is particularly suitable for parallel, vertical or angled parking situations.

Claims

Patent claims 1. Method for dynamically planning a parking trajectory of a vehicle (1 ) during an at least partially automated parking process, the method comprising the following steps: - detecting an environment of the vehicle (1) by at least one sensor device (S10), - determining and setting a target position (Z1) for the vehicle (1) based on the environmental data (S11) detected by the at least one sensor device, - Planning a parking trajectory based on the specified target position (Z1) (S12), - Recalculations of the target position (Z2, Z3, Zx) during the parking process based on environmental data updated during the parking process (S13), - comparing the difference between the specified target position (Z1) and a mean value calculated from at least a subset of the target positions (Z2, Z3, Zx) recalculated during the parking process, whereby the specified target position (Z1) is updated if the difference between the specified target position (Z1) and the mean value calculated exceeds a predetermined limit value (S14), - Dynamic replanning of the parking trajectory based on the updated target position (Z2, Z3, Zx) (S15).

2. Method according to claim 1, wherein the target positions (Z2, Z3, Zx) recalculated during the parking process are weighted according to timeliness when forming the mean value.

3. The method according to claim 1 or 2, wherein the specified target position comprises a viewpoint and an orientation of the vehicle relative to a spatially fixed coordinate system, wherein the updating of the target position may relate to the viewpoint and / or the orientation.

4. Method according to one of the preceding claims, wherein the mean value is formed from at least three newly calculated target positions which take into account a parking space which is as completely recorded as possible.

5. Method according to one of the preceding claims, wherein the last recalculated target position is selected as the updating target position.

6. Method according to one of the preceding claims, wherein during the parking process the target position (Z2, Z3, Zx) is recalculated in each measuring cycle of the at least one sensor device.

7. Method according to one of the preceding claims, wherein the parking operation is an angled parking operation or a vertical parking operation.

8. Method according to one of the preceding claims, wherein the speed of the vehicle (1) when driving through the predetermined parking trajectory is between 0.2 m / s and 3 m / s.

9. Parking assistance device for dynamically planning a parking trajectory of a vehicle (1) during an at least partially automatic parking process, with a sensor device for detecting an environment of the vehicle, an evaluation unit which is designed to determine and set a target position for the vehicle based on the environment data detected and transmitted by the at least one sensor device, wherein the evaluation unit is designed to plan a parking trajectory based on the set target position, wherein the evaluation unit is configured to recalculate target positions during the parking process based on the environmental data updated during the parking process, wherein the evaluation unit is configured to perform a difference comparison of the specified target position (Z1) with a calculated mean value from at least a subset of the target positions (Z2, Z3, Zx) recalculated during the parking process and to output an update of the specified target position if the difference between the specified target position (Z1) and the calculated mean value exceeds a predetermined limit value, wherein the evaluation unit is configured to perform dynamic replanning of the parking trajectory based on the updated target position (Z2, Z3, Zx). A motor vehicle with a parking assistance device according to claim 9.

Images