Method for planning a parking maneuver supported by a parking assistance system

Abstract

The invention relates to a method for planning a parking process of a vehicle (1) supported by a parking assistance system, the method comprising the following steps: - detecting at least one parking scenario in the surrounding area of ​​the vehicle (1) (S10); - planning a parking trajectory (PT) between a starting position (SP) and a target position (ZP) of the vehicle (1) by means of a first planning method, the first planning method being based on rule-based path planning for the detected parking scenario (S11); - checking whether at least a part of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) can be determined by means of the first planning method (S12); - if at least a part of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) of the vehicle (1) cannot be planned by means of the first planning method (S13), planning at least a part of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) of the vehicle (1) by means of a second general planning method.

Description

Technical Field

[0001] The present invention relates to a method for planning a parking maneuver supported by a parking assistance system, a parking assistance system, and a vehicle having a parking assistance system. Background Art

[0002] Known parking assistance systems from the prior art can be used to plan and implement a parking process in a semi-automatic or fully automatic manner. Here, the vehicle is maneuvered to the parking position by the parking assistance system on a parking trajectory calculated in advance in the planning phase.

[0003] It is known from the prior art to plan parking trajectories using deterministic planning methods that attempt to determine a parking trajectory based on calculation rules and geometric methods. These deterministic planning methods have one or more sets of calculation rules, each of which is associated with a specific parking scenario, such as reverse parking in a longitudinal parking situation. A disadvantage of these deterministic planning methods is their limited flexibility in the event of deviations from the basic parking scenario. They also lack the ability to react flexibly to changing parking situations (e.g., objects entering the parking trajectory), which often leads to failed parking attempts.

[0004] Furthermore, known general planning methods are not limited to specific parking scenarios but can be applied to a wide range of parking scenarios and can calculate parking trajectories even in complex, time-varying parking scenarios. A disadvantage of general planning methods is that they are computationally very complex, requiring high processor performance and a large amount of memory. Furthermore, even with minor changes in the parking situation, general planning methods produce completely different parking behaviors, which is disadvantageous to drivers, who expect relatively stable parking behavior from parking assistance systems. Furthermore, general planning methods struggle to adhere to the characteristics specified by each driver, as they cannot differentiate between parking scenarios and always use the same algorithm to calculate the parking trajectory, regardless of the parking situation. Summary of the Invention

[0005] Based on this, one object of the present invention is to provide a method for planning a parking maneuver supported by a parking assistance system, which method allows a time-saving and flexible calculation of the parking trajectory with a high driver acceptance and a low rate of failed parking attempts.

[0006] According to a first aspect, the present invention relates to a method for planning a parking maneuver supported by a parking assistance system. The method comprises the following steps:

[0007] First, at least one parking scenario is detected in the vehicle's surrounding area. The detected parking scenario is preferably classified, i.e., for example, it is determined which type of parking process is involved (forward parking / reverse parking, parallel parking / diagonal parking / perpendicular parking, etc.).

[0008] A parking trajectory between the vehicle's starting position and target position is then planned using a first planning method based on rule-based path planning for the detected parking scenario. The first planning method preferably uses a planning method suitable for the respective parking scenario type (forward parking, reverse parking, longitudinal parking, diagonal parking, perpendicular parking, etc.). The first planning method is preferably a deterministic planning method in which the parking trajectory is constructed using geometric curves using specialized calculation rules adapted to the detected parking scenario. The curves are preferably explicitly describable using mathematical functions.

[0009] After planning the parking trajectory using the first planning method, it is checked whether at least a portion of the parking trajectory between the starting position and the target position can be determined using the first planning method. In other words, an attempt is first made to plan at least a section of the entire trajectory using the first planning method.

[0010] If at least a portion of the parking trajectory between the vehicle's starting position and the target position cannot be planned using the first planning method, then at least a portion of the parking trajectory between the vehicle's starting position and the target position is planned using a second, general planning method. In other words, the remaining portion of the parking trajectory, or if no segment can be planned using the first planning method, the entire parking trajectory is planned using the second planning method. The second planning method is preferably a planning method that can freely plan the parking trajectory independent of the specific parking scenario. In other words, the second planning method is a universally applicable planning method that can be applied to a wide range of parking scenarios.

[0011] A major advantage of the method according to the present invention is that, by combining the first and second planning methods, a high computational efficiency is achieved while reducing the probability of failed planning attempts, since even in parking situations that cannot be planned using the first planning method from the outset or due to changes during the planning or implementation phase, the parking trajectory can be planned using the second general planning method.

[0012] According to one embodiment, the first planning method provides multiple sets of calculation rules, each of which is tailored to a corresponding identified parking scenario. Thus, using each set of calculation rules, at least a portion of the parking trajectory can be determined by combining multiple analytically well-defined trajectory segments. The set of calculation rules can, in particular, define geometric curves such as straight lines, circular arc segments, and / or clothoids, which, when strung together, form at least a portion of the parking trajectory.

[0013] According to one embodiment, a set of calculation rules is applied based on the detected parking scenario. Depending on the type of parking scenario detected (forward parking / reverse parking, longitudinal parking / diagonal parking / perpendicular parking, etc.), a set of calculation rules can be selected and used to attempt to determine a parking trajectory.

[0014] According to one embodiment, a first planning method performs segmented parking trajectory planning, comprising a plurality of parking trajectory segments, particularly based on geometric curves that are each analytically determined using one or more mathematical functions. The analytical determination of the geometric curves preferably incorporates surrounding objects identified in the respective parking situation (e.g., detected using surroundings recognition). This allows an attempt to analytically construct at least a portion of the parking trajectory, which advantageously saves computing power and achieves the desired parking maneuver in identical or substantially identical parking scenarios.

[0015] According to one embodiment, the geometric curve includes a straight line, a circular arc segment and / or a clothoid curve. Many common parking scenarios can be covered by these curves.

[0016] According to one embodiment, each calculation rule defines one or more parking trajectory segments. For example, a calculation rule may reflect the course of a geometric curve in two-dimensional space. Thus, a segmented parking trajectory can be analytically determined using the calculation rule set, where multiple parking trajectory segments are connected to calculate the parking trajectory.

[0017] According to one embodiment, a parking trajectory is planned segment by segment, with a first portion of the parking trajectory being determined using a first planning method and a second portion being determined using a second planning method. An attempt is preferably made to continue using at least some of the results of trajectory planning performed using the first planning method, for example, one or more calculated trajectory segments, and to use the second planning method to determine the remaining portion of the parking trajectory that could not be determined using the first planning method. This allows for savings in computing power even in more complex parking situations that cannot be fully calculated using the first planning method.

[0018] According to one embodiment, the second planning method utilizes a calculation algorithm that is independent of the detected planning scenario and / or encompasses several different parking scenarios. "Independent of the detected planning scenario" here means that the calculation algorithm is not tailored to a specific planning scenario, but rather freely plans a trajectory based on the identified vehicle surroundings. This allows for finding a suitable parking trajectory even in complex or changing parking situations.

[0019] According to one embodiment, the method further includes, as an additional step, at least semi-automatically performing the parking process using a parking assistance system. If the parking situation changes during the parking process, the parking trajectory between the vehicle's current position and the target position is replanned. More specifically, the planning is performed by attempting to determine at least a portion of the parking trajectory using a first planning method in a first step. If at least a portion of the parking trajectory between the vehicle's current position and the target position cannot be planned using the first planning method (i.e., the first planning method cannot determine the parking trajectory or can only determine a portion of the parking trajectory), planning the at least portion of the parking trajectory using a second planning method. This allows replanning to be performed during the parking process, i.e., after the parking trajectory has been planned, if the changed parking situation requires it.

[0020] According to one embodiment, a parking trajectory is planned from a vehicle's starting position to a planned target position, or vice versa. For example, the planning direction may be adjusted based on a recognized parking scene or a recognized surrounding area. By selecting the planning direction, at least a portion of the parking trajectory can be determined using a first planning method.

[0021] According to another aspect, the present invention relates to a parking assistance system for controlling a vehicle parking process. The parking assistance system comprises:

[0022] a detection unit for detecting at least one parking scene in the vehicle's surrounding area;

[0023] a computing unit configured to plan a parking trajectory between a starting position and a target position of the vehicle using a first planning method, wherein the first planning method is based on rule-based path planning for a detected parking scenario;

[0024] a computing unit configured to check whether at least a portion of the parking trajectory between the starting position and the target position can be determined by a first planning method; and

[0025] a computing unit configured to plan at least a portion of the parking trajectory between the starting position and the target position of the vehicle using a second general planning method if at least a portion of the parking trajectory between the starting position and the target position of the vehicle cannot be planned using the first planning method.

[0026] According to another aspect, the present invention relates to a vehicle having a parking assistance system as described above.

[0027] For the purposes of the present invention, a "parking process" refers to the process of maneuvering a vehicle forward, reverse, sideways, or diagonally into a parking space. The driver can be inside or outside the vehicle during the parking process (remote control parking process). A "parking process" in this context can include the process of entering a parking space or the process of leaving a parking space.

[0028] In the sense of the present invention, a “parking assistance system” is understood to be any system which supports the driver during the parking process and enables at least a semi-automatic parking process.

[0029] In the meaning of the present invention, the terms “approximately”, “substantially” or “approximately” mean that the deviation from the respective exact value is + / - 10%, preferably + / - 5%, and / or the deviation is a variation that is insignificant to the function.

[0030] Further embodiments, advantages, and possible applications of the present invention are also apparent from the following description of exemplary embodiments and the accompanying drawings. All described and / or illustrated features, alone or in any combination, constitute the subject matter of the present invention, regardless of their combination in the claims or their references thereto. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] The present invention will be described in detail below with reference to the accompanying drawings of the embodiments, wherein:

[0032] Figure 1 A parking situation with an unoccupied parking space in a longitudinal parking situation is shown by way of example and diagrammatic representation;

[0033] Figure 2 Shown in an exemplary and schematic manner Figure 1 A similar parking situation, but with an object that makes the planning of the parking scenario more complicated.

[0034] Figure 3 A block diagram is shown by way of example to illustrate the process of a method for planning a parking maneuver supported by a parking assistance system; and

[0035] Figure 4 A flow chart for planning a parking maneuver supported by a parking assistance system using a first planning method and a second planning method is presented by way of example. DETAILED DESCRIPTION

[0036] Figure 1 A vehicle 1 is shown in the context of a parallel parking situation in which there is an unoccupied parking space P1 between two vehicles. Vehicle 1 has a parking assistance system, by means of which vehicle 1 can be parked at least semi-automatically in unoccupied parking space P1.

[0037] During the parking process, the vehicle 1 is guided by the parking assistance system along a parking trajectory PT. Figure 1 A parking trajectory PT is shown as an example. A parking trajectory PT can consist of several segments. For example, a segment can consist of a straight line or a circular arc segment. Furthermore, to further improve the comfort of the parking assistance system, a segment can also consist of a clothoid. The clothoid can compensate for or offset changes in curvature between two circular arc segments or between a straight line and a circular arc segment.

[0038] In the illustrated embodiment, the parking trajectory includes five segments, wherein the following segments are used from the starting position SP of the vehicle 1 to the target position ZP, for example: the first segment S1 is a straight line, the second segment S2 is a circular arc segment, the third segment S3 is a clothoid, the fourth segment S4 is a circular arc segment, and the fifth segment S5 is a straight line.

[0039] This type of parking situation can be fully addressed using a first, rule-based planning method that uses geometric methods to determine the parking trajectory. Depending on the identified parking situation (parallel parking, parallel parking, diagonal parking, etc.), the first planning method uses a deterministic algorithm consisting of several sets of calculation rules, each of which is tailored to the identified parking situation. This deterministic approach ensures that defined and reproducible conditions always occur in each parking scenario, allowing the parking assistance system to provide stable and predictable parking behavior.

[0040] and Figure 1 Compared with the parking situation in Figure 2 A slightly modified parking situation is shown, in which an object O forces a change in the parking trajectory PT. For example, the object O may have been present before the parking trajectory (PT) was first planned, or it may have unexpectedly appeared in the parking situation (e.g., an open car door) and thus necessitate a replanning of the parking trajectory (PT) (after the initial planning).

[0041] The object O changes the parking situation in such a way that a suitable parking trajectory PT cannot be completely determined at least by the first planning method. In other words, no set of calculation rules can be found by which a suitable parking trajectory PT can be calculated in a deterministic manner.

[0042] If the parking trajectory PT cannot be calculated using the deterministic first planning method, at least a portion of the parking trajectory PT can be calculated using a second planning method based on a general planning model. In other words, the second planning method does not perform path planning specific to each parking scenario. Instead, the second planning method attempts to determine the parking trajectory PT based on each identified situation using a general algorithm that can be applied to a number of different parking scenarios.

[0043] The entire parking trajectory PT can be calculated using the second planning method, or both planning methods can be used to calculate the entire parking trajectory PT by calculating a first part T1 of the parking trajectory PT using a first rule-based planning method and a second part T2 of the parking trajectory PT using a second universal planning method.

[0044] Figure 2 An example of a parking trajectory PT that can be determined using a combination of a first planning method and a second planning method is presented. The parking trajectory PT is planned segment by segment, specifically in the second segment between the starting position SP and the intermediate point A (corresponding to the second portion T2) and in the first segment between the intermediate point A and the target position ZP (corresponding to the first portion T1). A generic second planning method can be used for the second segment, which allows for navigating around the object O, while a deterministic first planning method can be used for the first segment. The combination of these two planning methods advantageously allows the parking assistance system to provide a parking maneuver that is easier for the respective vehicle driver to understand, since the parking assistance system attempts to calculate the parking trajectory PT based on the first, rule-based planning method as much as possible.

[0045] Figure 3 A block diagram of a method according to the present invention for planning a vehicle parking process supported by a parking assistance system is shown.

[0046] In a first step ( S10 ), a parking scenario is detected in the vehicle's surrounding area. In particular, the type of parking scenario involved (parallel parking, parallel parking, diagonal parking, etc.) is determined. For example, several different parking scenario variants can be classified in the parking assistance system of vehicle 1 , and an attempt can be made to assign the detected parking scenario to one of the parking scenario variants.

[0047] Subsequently, in a further step ( S11 ), a parking trajectory PT between the starting position SP and the target position ZP of the vehicle 1 is planned or calculated using a first planning method based on rule-based path planning for the detected parking scenario. This planning can in particular be a planning attempt.

[0048] It can then be checked whether at least a portion of the parking trajectory PT between the starting position SP and the target position ZP can be determined by the first planning method ( Figure 2 In other words, it can be determined whether the parking trajectory PT can be calculated completely or at least partially based on the first planning method.

[0049] If at least a portion of the parking trajectory PT between the vehicle 1 starting position SP and the target position ZP ( Figure 2 If the portion T2 in the plan cannot be planned using the first planning method, at least a portion of the parking trajectory PT between the starting position SP and the target position ZP of the vehicle 1 can be calculated using a universal second planning method ( S13 ).

[0050] Figure 4 A flow chart is shown for calculating a parking trajectory PT based on the method according to the present invention.

[0051] As described above, each corresponding current parking scene is first detected ( S20 ).

[0052] Subsequently, the parking trajectory PT is established using the calculation rules of the first planning method (S21). Here, the calculation rules can be selectively selected for each corresponding parking scenario.

[0053] Subsequently, in step S22 , it is checked whether the entire parking trajectory PT can be planned with the aid of the first planning method, ie whether a parking trajectory PT from the starting position SP to the target position ZP of the vehicle 1 can be found.

[0054] If the entire parking trajectory PT can be planned with the aid of the first planning method, the method ends ( S23 ).

[0055] However, if the entire parking trajectory PT cannot be planned using the first planning method, a check is performed to see whether a portion of the parking trajectory PT planned using the first planning method can continue to be used (S24). Figure 2 In the example shown, this refers to, for example, the portion T1 of the parking trajectory. Thus, for example, the second general planning method can be used to calculate the portion of the parking trajectory PT by which the object O is driven around, while the other portion, in particular the portion after the object, can be calculated using the first rule-based planning method.

[0056] After checking the usable parking trajectory segments calculated by the first planning method, the remaining portion of the parking trajectory PT (i.e., the remaining unplanned portion, or if no parking trajectory segments are still usable, the entire parking trajectory) can be planned using the second planning method ( S25 ).

[0057] The parking trajectory PT from the starting position SP to the target position ZP, or in the opposite direction, ie from the target position ZP to the starting position SP, can be planned depending on the parking situation and / or parking scenario.

[0058] The present invention has been described above with reference to embodiments thereof. It is noted that numerous modifications and variations are possible without departing from the scope of protection defined by the appended claims.

[0059] List of reference numerals:

[0060] 1 vehicle

[0061] A midpoint

[0062] PT parking track

[0063] S1, S2, S3 sections

[0064] SP starting position

[0065] ZP target position

Claims

1. A method for planning a parking maneuver for a vehicle (1) supported by a parking assistance system, the method comprising the following steps: - detecting at least one parking scenario (S10) in the surrounding area of ​​the vehicle (1); - planning a parking trajectory (PT) between a starting position (SP) and a target position (ZP) of the vehicle (1) by means of a first planning method, the first planning method being based on rule-based path planning for a detected parking scenario (S11); - checking whether at least a portion of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) can be determined by means of the first planning method (S12); - if at least a part of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) of the vehicle (1) cannot be planned by means of the first planning method (S13), planning at least a part of the parking trajectory (PT) between the starting position (SP) and the target position (ZP) of the vehicle (1) by means of a second universal planning method, The first planning method is based on geometric curves which are determined analytically by one or more mathematical functions. The calculation algorithm used in the second planning method is independent of the detected planning scenario and / or includes a plurality of different parking scenarios. The parking trajectory is planned segment by segment, wherein a first part (T1) of the parking trajectory (PT) is determined by a first planning method and a second part (T2) of the parking trajectory (PT) which cannot be determined by the first planning method is determined by a second planning method.

2. The method according to claim 1, characterized in that The first planning method provides multiple sets of calculation rules, wherein each set of calculation rules is tailored to a corresponding identified parking scenario.

3. The method according to claim 2, characterized in that Based on the detected parking scenario, a set of calculation rules are applied.

4. The method according to claim 1, characterized in that: The first planning method performs a segmented parking trajectory planning comprising a plurality of parking trajectory segments.

5. The method according to claim 4, characterized in that The geometric curve includes a straight line, a circular arc segment and / or a clothoid.

6. The method according to claim 4 or 5, characterized in that: Each calculation rule determines one or more parking trajectory segments.

7. The method according to claim 1, characterized in that As a further step, the method comprises: carrying out a parking process at least semi-automatically with the aid of a parking assistance system, and replanning a parking trajectory (PT) between a current position of the vehicle and a target position if a parking scenario change occurs during the parking process, wherein the replanning is carried out in the following manner: in a first step, an attempt is made to determine at least a portion of the parking trajectory based on a first planning method, and if at least a portion of the parking trajectory between the current position of the vehicle (1) and the target position (ZP) cannot be planned with the aid of the first parking method, at least a portion of the parking trajectory (PT) is planned with the aid of a second planning method.

8. The method according to claim 1, characterized in that A parking trajectory (PT) is planned from a starting position (SP) of a vehicle (1) to a planned target position (ZP) or from a planned target position (ZP) of the vehicle (1) to the starting position (SP).

9. A parking assistance system for a vehicle (1), the parking assistance system comprising: - a detection unit for detecting at least one parking scene in the surrounding area of ​​the vehicle (1); a computing unit configured to plan a parking trajectory (PT) between a starting position (SP) and a target position (ZP) of the vehicle (1) by means of a first planning method, wherein the first planning method is based on rule-based path planning for a detected parking scenario; a computing unit configured to check whether at least a portion of a parking trajectory (PT) between a starting position (SP) and a target position (ZP) can be determined by means of a first planning method; and a computing unit configured to plan at least a portion of a parking trajectory (PT) between a starting position (SP) and a target position (ZP) of the vehicle (1) by means of a second, universal planning method, if at least a portion of the parking trajectory (PT) between a starting position (SP) and a target position (ZP) of the vehicle (1) cannot be planned by means of the first planning method, The first planning method is based on geometric curves which are determined analytically by one or more mathematical functions. The calculation algorithm used in the second planning method is independent of the detected planning scenario and / or includes a plurality of different parking scenarios. The parking trajectory is planned segment by segment, wherein a first part (T1) of the parking trajectory (PT) is determined by a first planning method and a second part (T2) of the parking trajectory (PT) which cannot be determined by the first planning method is determined by a second planning method.

10. A vehicle comprising a parking assistance system according to claim 9.

Images