A trajectory planning method and device for combining multiple parking APA modes

Abstract

The present invention provides a trajectory planning method and device in the form of a combination of multiple parking APA modes, which solves the problems existing in the prior art in parking path planning, such as high computing power consumption, low planning success rate, long planning time and poor planning effect. The current parking scene type is determined by sensing and acquiring information around the vehicle, and then the final path planning route is generated through a parking planning method suitable for the current parking scene. The planning adaptability in the APA scene is improved, and the success rate of planning solutions in different scenes is greatly improved. At the same time, the consumption of the solution algorithm is effectively reduced, and the solution speed is improved.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a trajectory planning method and device for a parking APA in the form of a combination of multiple modes. Background Art

[0002] APA (Automatic Parking Assist) is an advanced automotive technology designed to help drivers complete parking operations more easily and safely. This system usually uses sensors on the vehicle (such as ultrasonic sensors or cameras) to detect parking spaces and automatically controls the steering wheel to guide the vehicle into the parking space.

[0003] Existing parking planning algorithms can be roughly divided into pure geometric method, pure search method, pure sampling method and numerical optimization method; for pure geometric method, the advantage is simple implementation and low computing power consumption; the disadvantage is that the algorithm has poor versatility and cannot cover all planning scenarios. For slightly more complex scenarios, the success rate of trajectory solution is greatly reduced; for pure search method and pure sampling method, the advantage is high flexibility and strong versatility; the disadvantage is high computational complexity, high computing power consumption and time-consuming search; for numerical optimization method, the advantage is that various constraints can be added and high-quality paths can be output; the disadvantage is high computational complexity, sensitivity to initial solution, and may fall into local optimal problems. Summary of the invention

[0004] The present invention provides a trajectory planning method and device in the form of a combination of multiple parking APA modes, so as to solve the problems existing in the prior art in parking path planning, such as high computing power consumption, low planning success rate, long planning time and poor planning effect.

[0005] In a first aspect, the present invention provides a trajectory planning method in the form of a combination of multiple parking APA modes, which specifically includes the following steps:

[0006] Step S1, obtaining information about obstacles around the vehicle, parking space information (i.e., planning end point) and the current position information of the vehicle, taking the current position information of the vehicle as the planning starting point, and taking the parking space information as the planning end point;

[0007] Step S2: determining the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle;

[0008] Step S3: determining a parking planning method according to the parking space scene type;

[0009] Step S4: planning a parking path according to the planned starting point, the planned end point and the determined parking planning method to form a parking planning route.

[0010] Preferably, in step S2, the parking space scene types include: vertical parking space, inclined parking space and horizontal parking space;

[0011] The vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces;

[0012] Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

[0013] Preferably, in step S3, the parking planning method includes two methods; wherein the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm; the second parking planning method includes the search method Hybrid A Star algorithm, the Target Tree algorithm, the SpaceInside algorithm and the StraightLine algorithm;

[0014] Among them, for vertical parking spaces, inclined parking spaces, horizontal line parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning.

[0015] Preferably, when performing parking path planning by the first parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm.

[0016] Preferably, when the parking path is planned by the second parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm.

[0017] Preferably, if the current parking space scene type is a horizontal double-boundary parking space scene, the planning of the parking path in step S4 specifically includes the following steps:

[0018] Step S401a, determining the coordinates of five path planning points of the current horizontal double-boundary parking space scene, which are the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point, wherein the first path planning point is the planning starting point and the fifth path planning point is the planning end point;

[0019] Step S401b, performing path planning on the first path planning point and the second path planning point by using the Hybrid A Star algorithm to form a first path;

[0020] Step S401c, performing path planning on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path;

[0021] Step S401d: performing path planning on the third path planning point and the fourth path planning point by using the SpaceInside algorithm to form a third path;

[0022] Step S401e, performing path planning on the fourth path planning point and the fifth path planning point by using the StraightLine algorithm to form a fourth path;

[0023] Step S401f: sequentially combine the first path, the second path, the third path and the fourth path to generate a complete parking planning route.

[0024] Preferably, if the current parking space scene type is a non-horizontal double-boundary parking space scene, the planning of the parking path in step S4 specifically includes the following steps:

[0025] Step S402a, determining the coordinates of three path planning points of the current non-horizontal double-boundary parking space scene, which are the sixth path planning point, the seventh path planning point and the eighth path planning point, wherein the sixth path planning point is the planning starting point and the eighth path planning point is the planning end point;

[0026] Step S402b, performing path planning on the sixth path planning point and the seventh path planning point by using the Hybrid A Star algorithm to form a fifth path;

[0027] Step S402c, performing path planning on the seventh path planning point and the eighth path planning point by using the StraightLine algorithm to form a sixth path;

[0028] Step S402d: Combine the fifth path and the sixth path to generate a complete parking planning route.

[0029] Preferably, in step S401c, path planning is performed on the second path planning point and the third path planning point by using a TargetTree algorithm to form a second path, which specifically includes the following steps:

[0030] Step S401c1, constructing a continuous curvature target tree according to the second path planning points to form a target tree set;

[0031] Step S401c2: establishing an optimal target tree cost function according to the target tree set, and selecting the optimal target tree therein;

[0032] Step S401c3: establishing an optimal node cost function according to the optimal target tree, and selecting an optimal node (i.e., the third path planning point);

[0033] Step S401c4: According to the optimal node, a path between the path planning starting point and the optimal node is planned by a path planning algorithm.

[0034] Preferably, in step S401d, the third path planning point and the fourth path planning point are path planned by SpaceInside algorithm to form a third path; wherein the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking exit direction; parking in is considered from the perspective of parking exit;

[0035] Among them, ACBD, bdec and FF'G'G are all rectangular areas and the four vertices of each rectangle correspond to each other (in the present application, the mutual correspondence means that point A corresponds to point b, point A corresponds to point F, point b corresponds to point F, and at the same time, line segment FF' is parallel to line segment AC and line segment bd); among them, ACBD represents a parking space; bdec represents the area where the vehicle is located; FF'G'G represents the boundary abstracted from the obstacle; in the rectangular area FF'G'G, point F and point G are connected through point F' and point G' in turn; in the rectangular area ACBD, point A and point B are connected through point C and point D in turn; the width of the rectangular area FF'G'G is FF' (the length of the line segment FF' is L FF' ), the length is F'G' (the length of line segment F'G' is L F'G' ), the width of the rectangular area ACBD is AC (the length of the line segment AC is L AC ), the length is CD (the length of line segment CD is L CD ), the width of bdec is bd (the length of line segment bd is L bd ), length is de (the length of line segment de is L de ), and L FF' >L AC >L bd, L F'G' >L CD >L de .

[0036] Among them, line segment ec represents the front end of the vehicle, line segment bd

[0037] Indicates the rear end of a vehicle.

[0038] Preferably, in step S401d, considering parking in from the perspective of parking out (i.e., taking the fourth path planning point as the planning starting point of the SpaceInside algorithm and the third path planning point as the planning end point of the SpaceInside algorithm to perform path planning) specifically includes the following steps:

[0039] Step S401d1, the midpoint of line segment CD is set to O, the positive direction of the Y axis is the direction passing through point O and the same as the direction of vector OC, and the positive direction of the X axis is the direction passing through point O and the same as the direction of vector CA, and an XOY coordinate system is constructed;

[0040] Step S401d2, initializing line segment ec to collide with line segment GG' as the initial position of the vehicle, wherein the rear axle axis of the vehicle is O1;

[0041] Step S401d3, the vehicle coordinate system is constructed by passing through the intersection of point O1 and line segment ce and parallel to de as the positive direction of the Y1 axis, and passing through the intersection of point O1 and line segment de and parallel to bd as the positive direction of the X1 axis;

[0042] Step S401d4, after the vehicle steering wheel is rotated to the extreme position (i.e., the steering wheel is turned to the end) in the first direction (in this application, the first direction refers to the rotation direction in which the rear end of the vehicle approaches the line segment F'G' during the vehicle backing up), the vehicle performs a backing operation, and the steering angle when the vehicle point b contacts the line segment FF' is determined. The steering angle when the vehicle point d contacts the line segment F'G' ,Pick ;

[0043] Step S401d5: when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rearward rotation center of the vehicle is calculated;

[0044] Step S401d6: According to the , the backward rotation center calculates the backward motion trajectory of the vehicle and the backward motion endpoint;

[0045] Step S401d7, set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the vehicle steering wheel in the second direction (in this application, the second direction refers to the rotation direction that makes the rear end of the vehicle move away from F'G' during the forward movement of the vehicle) to the extreme position, and then the vehicle moves forward to determine the steering angle when the vehicle point d contacts the line segment F'G' The steering angle at which the vehicle point e contacts the line segment GG' ,Pick ;

[0046] Step S401d8, when the vehicle contacts the boundary line segment of area FF'G'G, calculate the forward rotation center of the vehicle, otherwise, take the X-axis coordinate value of the vehicle point e on the trajectory that is the same as or slightly higher than the X-axis coordinate value of point G, and output the motion trajectory of the SpaceInside algorithm;

[0047] Step S401d9: According to the , the forward rotation center calculates the forward phase motion trajectory of the vehicle and the forward phase motion end point;

[0048] Step S401d10: Repeat steps S401d4 to S401d9 until the vehicle does not come into contact with the boundary line segment of area FF'G'G, and then output the motion trajectory of the SpaceInside algorithm.

[0049] Preferably, in step S401d4, when the steering wheel of the vehicle is rotated to the extreme position in the first direction for the first time, point b contacts the obstacle area before point d, which is optimal so that the next section of the kneading track can adjust the heading to the maximum extent.

[0050] More preferably, in order to ensure that the moving point b of the vehicle first contacts the obstacle area FF'G'G in the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows:

[0051]

[0052] in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the Y-axis coordinate value of point b when the vehicle first reaches the end point of the backward movement.

[0053] Preferably, in step S401d5, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rotation center of the vehicle is calculated as follows:

[0054]

[0055] Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinates of the vehicle in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows:

[0056]

[0057] Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows:

[0058]

[0059] in, Indicates the angle of rotation. When the rear end of the vehicle rotates toward the line segment F'G', the rotation angle is positive, negative, rotation angle is negative.

[0060] Preferably, after the StraightLine algorithm generates a straight line between the seventh path planning point and the eighth path planning point, the straight line is discretized into a series of path points at equal intervals to obtain the path output by the StraightLine algorithm. The purpose of path planning by the StraightLine algorithm is to leave room for adjustment for subsequent control. If the previous tracking error is large, the StraightLine algorithm can be used to correct the position of the vehicle.

[0061] In a second aspect, the present invention further provides a trajectory planning device in the form of a combination of multiple parking APA modes, which specifically includes the following modules:

[0062] The information acquisition module is used to obtain the obstacle information around the vehicle, the parking space information and the current position information of the vehicle. The current position information of the vehicle is the planning starting point, and the parking space information is the planning end point;

[0063] A parking space scene determination module, used to determine the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle;

[0064] A parking planning method determination module, used to determine a parking planning method according to the parking space scene type;

[0065] The parking route planning module is used to plan a parking path according to the planning starting point, the planning end point and the determined parking planning method to form a parking planning route.

[0066] Preferably, in the parking space scene determination module, the parking space scene types include: vertical parking space, inclined parking space and horizontal parking space;

[0067] The vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces;

[0068] Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

[0069] Preferably, in the parking planning method determination module, the parking planning methods include two types; wherein the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm; the second parking planning method includes the search method Hybrid A Star algorithm, the Target Tree algorithm, the SpaceInside algorithm and the StraightLine algorithm;

[0070] Among them, for vertical parking spaces, inclined parking spaces, horizontal line parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning.

[0071] Preferably, when performing parking path planning by the first parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm.

[0072] Preferably, when the parking path is planned by the second parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm.

[0073] Preferably, if the current parking space scene type is a horizontal double-boundary parking space scene, the parking route planning module specifically includes the following submodules:

[0074] The first submodule of parking route planning is used to determine the coordinates of five path planning points of the current horizontal double-boundary parking space scene, namely the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point, wherein the first path planning starting point is the planning starting point and the fifth path planning point is the planning end point;

[0075] The second submodule of parking route planning is used to perform path planning on the first path planning point and the second path planning point by using the Hybrid A Star algorithm to form a first path;

[0076] The third submodule of parking route planning is used to perform path planning on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path;

[0077] The fourth submodule of parking route planning is used to perform path planning on the third path planning point and the fourth path planning point by using the SpaceInside algorithm to form a third path;

[0078] The fifth submodule of parking route planning is used to perform path planning on the fourth path planning point and the fifth path planning point by using the StraightLine algorithm to form a fourth path;

[0079] The sixth submodule of parking route planning is used to combine the first path, the second path, the third path and the fourth path to generate a complete parking planning route.

[0080] Preferably, if the current parking scene type is a non-horizontal double-boundary parking scene, the parking route planning module specifically includes the following submodules:

[0081] The seventh submodule of parking route planning is used to determine the coordinates of three path planning points of the current non-horizontal double-boundary parking space scene, which are respectively recorded as the sixth path planning point, the seventh path planning point and the eighth path planning point, wherein the sixth path planning point is the planning starting point and the eighth path planning point is the planning end point;

[0082] The eighth submodule of parking route planning is used to perform path planning on the sixth path planning point and the seventh path planning point by using the Hybrid A Star algorithm to form a fifth path;

[0083] The ninth submodule of parking route planning is used to perform path planning on the seventh path planning point and the eighth path planning point by using the StraightLine algorithm to form a sixth path;

[0084] The tenth submodule of parking route planning is used to combine the fifth path and the sixth path to generate a complete parking planning route.

[0085] Preferably, the third submodule of parking route planning specifically includes the following submodules:

[0086] A first grandchild module, configured to construct a continuous curvature target tree according to the second path planning points to form a target tree set;

[0087] The second grandchild module is used to establish an optimal target tree cost function according to the target tree set, and select the optimal target tree therein;

[0088] The third grandchild module is used to establish an optimal node cost function according to the optimal target tree and select an optimal node (i.e., a third path planning point);

[0089] The fourth submodule is used to plan the path between the path planning starting point and the optimal node through a path planning algorithm according to the optimal node.

[0090] Preferably, in the fourth submodule of parking route planning, the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking direction; parking is considered from the perspective of parking.

[0091] Among them, ACBD, bdec and FF'G'G are all rectangular areas and the four vertices of each rectangle correspond to each other; among them, ACBD represents the parking space; bdec represents the area where the vehicle is located; FF'G'G represents the boundary abstracted from the obstacle; in the rectangular area FF'G'G, point F and point G are connected through point F' and point G' in turn; in the rectangular area ACBD, point A and point B are connected through point C and point D in turn; the width of the rectangular area FF'G'G is FF' (the length of the line segment FF' is L FF'), the length is F'G' (the length of line segment F'G' is L F'G' ), the width of the rectangular area ACBD is AC (the length of the line segment AC is L AC ), the length is CD (the length of line segment CD is L CD ), the width of bdec is bd (the length of line segment bd is L bd ), length is de (the length of line segment de is L de ), and L FF' >L AC >L bd , L F'G' >L CD >L de .

[0092] The line segment ec represents the front end of the vehicle, and the line segment bd represents the rear end of the vehicle.

[0093] Preferably, the fourth submodule of parking route planning specifically includes the following submodules:

[0094] The fifth module is used to set the midpoint of line segment CD to O, the line passing through point O and in the same direction as vector OC is the positive direction of the Y axis, and the line passing through point O and in the same direction as vector CA is the positive direction of the X axis, to construct an XOY coordinate system;

[0095] The sixth submodule is used to initialize the line segment ec to collide with the line segment GG' as the initial position of the vehicle, wherein the rear axle axis of the vehicle is O1;

[0096] The seventh grandson module is used to construct the vehicle coordinate system by passing through the intersection of point O1 and line segment ce and parallel to de as the positive direction of Y1 axis, passing through the intersection of point O1 and line segment de and parallel to bd as the positive direction of X1 axis;

[0097] The eighth submodule is used to determine the steering angle when the vehicle point b contacts the line segment FF' after the vehicle steering wheel is rotated to the extreme position in the first direction (i.e. the steering wheel is turned to the limit) and the vehicle performs a reverse operation The steering angle when the vehicle point d contacts the line segment F'G' ,Pick ;

[0098] The ninth submodule is used to calculate the vehicle's backward rotation center when the vehicle contacts the boundary line segment of the rectangular area FF'G'G;

[0099] The tenth grandchild module is used according to the , the backward rotation center calculates the backward motion trajectory of the vehicle and the backward motion endpoint;

[0100] The eleventh module is used to set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the vehicle steering wheel in the second direction to the extreme position, and then the vehicle moves forward to determine the steering angle when the vehicle point d contacts the line segment F'G'. The steering angle at which the vehicle point e contacts the line segment GG' ,Pick ;

[0101] The twelfth module is used to calculate the forward rotation center of the vehicle when the vehicle contacts the boundary line segment of the area FF'G'G. Otherwise, the X-axis coordinate value of the vehicle point e on the trajectory is the same as or slightly higher than the X-axis coordinate value of point G, and output the motion trajectory of the SpaceInside algorithm;

[0102] The thirteenth grandson module is used according to the , the forward rotation center calculates the forward phase motion trajectory of the vehicle and the forward phase motion end point;

[0103] The fourteenth grandson module is used to repeatedly execute the fifth grandson module, the sixth grandson module, the seventh grandson module, the eighth grandson module, the ninth grandson module, the tenth grandson module, the eleventh grandson module, the twelfth grandson module and the thirteenth grandson module until the vehicle does not contact the boundary segment of the area FF'G'G, and output the motion trajectory of the SpaceInside algorithm.

[0104] Preferably, in the eighth submodule, when the steering wheel of the vehicle rotates to the extreme position in the first direction for the first time, point b contacts the obstacle area before point d, which is optimal so that the next section of the kneading track can adjust the heading to the maximum extent.

[0105] More preferably, in order to ensure that the moving point b of the vehicle first contacts the obstacle area FF'G'G in the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows:

[0106]

[0107] in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the Y-axis coordinate value of point b when the vehicle first reaches the end point of the backward movement.

[0108] Preferably, in the ninth submodule, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rotation center of the vehicle is calculated as follows:

[0109]

[0110] Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinates of the vehicle in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows:

[0111]

[0112] Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows:

[0113]

[0114] in, Indicates the angle of rotation. When the rear end of the vehicle rotates toward the line segment F'G', the rotation angle is positive, negative, rotation angle is negative.

[0115] Preferably, after the StraightLine algorithm generates a straight line between the seventh path planning point and the eighth path planning point, the straight line is discretized into a series of path points at equal intervals to obtain the path output by the StraightLine algorithm. The purpose of path planning by the StraightLine algorithm is to leave room for adjustment for subsequent control. If the previous tracking error is large, the StraightLine algorithm can be used to correct the position of the vehicle.

[0116] In a third aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a trajectory planning method in the form of a combination of multiple parking APA modes as described in any one of the first aspect of the present application.

[0117] In a fourth aspect, the present invention further provides an electronic device, comprising: a memory storing a computer program; and a processor communicating with the memory, for executing a trajectory planning method in the form of a combination of multiple parking APA modes described in any one of the first aspects of the present application when the computer program is called.

[0118] Compared with the prior art, the present invention has the following obvious outstanding substantive features and significant advantages:

[0119] The present invention provides a trajectory planning method and device in the form of a combination of multiple parking APA modes, which solves the problems existing in the prior art in parking path planning, such as high computing power consumption, low planning success rate, long planning time and poor planning effect. The current parking scene type is determined by sensing and acquiring information around the vehicle, and then the final path planning route is generated through a parking planning method suitable for the current parking scene. The planning adaptability in the APA scene is improved, and the success rate of planning solutions in different scenes is greatly improved. At the same time, the consumption of the solution algorithm is effectively reduced, and the solution speed is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0120] The accompanying drawings, which constitute part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention. In the accompanying drawings:

[0121] Figure 1 It is a flow chart of a trajectory planning method in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0122] Figure 2 It is a schematic diagram of constructing a continuous curvature target tree in the TargetTree algorithm in a trajectory planning method in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0123] Figure 3 It is a schematic diagram of a vehicle being located at a planned end point in a trajectory planning method in a parking APA multiple mode combination form according to a preferred embodiment of the present invention.

[0124] Figure 4 It is a schematic diagram of the reverse planning starting point of the SpaceInside algorithm in a trajectory planning method in a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0125] Figure 5 It is a schematic diagram of vehicle rotation using the SpaceInside algorithm in a trajectory planning method in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0126] Figure 6 It is a schematic diagram of a vehicle rotating around a rotation center in a trajectory planning method in a parking APA multiple mode combination form according to a preferred embodiment of the present invention.

[0127] Figure 7 It is a schematic diagram of the calculation of the vehicle rotation angle in a trajectory planning method in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0128] Figure 8It is a parking schematic diagram of the SpaceInside algorithm in a trajectory planning method of a parking APA multiple mode combination in a preferred embodiment of the present invention.

[0129] Fig. 9 It is a parking planning schematic diagram of a trajectory planning method in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention.

[0130] Fig.10 It is a schematic diagram of the structure of a trajectory planning device in the form of a combination of multiple parking APA modes in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0131] The present invention provides a trajectory planning method and device in the form of a combination of multiple parking APA modes. In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0132] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0133] Embodiment 1:

[0134] like Figure 1-Figure 9 As shown, the trajectory planning method of a parking APA multiple modes combination method described in this embodiment specifically includes the following steps:

[0135] Step S1, obtaining information about obstacles around the vehicle, parking space information (i.e., planned end point), and current position information of the vehicle (i.e., planned starting point).

[0136] Step S2: determining the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle; wherein the parking space scene types include: vertical parking space, inclined parking space and horizontal parking space;

[0137] Among them, the vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces.

[0138] Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

[0139] Step S3: Determine a parking planning method according to the parking space scene type.

[0140] Among them, the parking planning methods include two types; among them, the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm, and the second parking planning method includes the search method Hybrid A Star algorithm, the Target Tree algorithm, the SpaceInside algorithm and the StraightLine algorithm.

[0141] Among them, for vertical parking spaces, inclined parking spaces, horizontal line parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning.

[0142] Among them, when planning the parking path through the first parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; at this time, the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm.

[0143] Among them, when planning the parking path through the second parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; at this time, the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm.

[0144] The main planning algorithms for the above two parking path planning methods are both the Habrid A Star algorithm. However, on the low computing power platform installed in the vehicle, the Habrid A Star algorithm is very likely to fail to plan a path within the specified time. Therefore, for the case of horizontal double-boundary parking spaces, a combination of four algorithms is used for parking path planning. For the case of non-horizontal double-boundary parking spaces, a combination of two algorithms is used for parking path planning. The StraightLine algorithm is to leave room for adjustment for the control. At the same time, if the previous tracking error is large, it can be corrected through the straight line segment here.

[0145] Step S4: planning a parking path according to the planned starting point, the planned end point and the determined parking planning method to form a parking planning route.

[0146] Optionally, if the current parking scene type is a horizontal double-boundary parking scene, step S4 specifically includes the following steps:

[0147] Step S401a, determine the coordinates of the five path planning points of the current horizontal double-boundary parking space scene, which are the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point; wherein the first path planning starting point is the planning starting point, and the fifth path planning point is the planning end point.

[0148] Step S401b: perform path planning on the first path planning point and the second path planning point by using the Hybrid A Star algorithm to form a first path. The Hybrid A Star algorithm specifically includes the following steps:

[0149] Step 1: In the initial state, define the starting node and the target node, and determine the target of path planning.

[0150] Step 2: discretize the continuous path planning space into multiple grid units, and assign a cost value to each grid unit by rasterizing the obstacle information of the environment; wherein the cost value is determined based on whether the grid unit is an obstacle and the density of the obstacle.

[0151] Step 3: Use the heuristic function to calculate the estimated cost from the current node to the target node. The heuristic function comprehensively considers the Euclidean distance and the Manhattan distance to improve the efficiency of path search.

[0152] Step 4: Use the A* search algorithm to search for paths in the grid mapping space. During the search, use a priority queue to store the nodes to be expanded, and sort the nodes according to their total cost (including the cost of the path that has been traveled and the estimated cost).

[0153] Step 5: Vehicle motion model within the grid: The vehicle motion model is introduced to calculate the steering and motion costs between nodes to ensure that the generated path is feasible in actual motion. The model takes into account physical constraints such as the minimum turning radius and maximum steering angle of the vehicle.

[0154] Step 6: If there is a Reeds-Shepp curve or a Hybrid-Curvature curve between the current node and the target node, and the curve is valid, the validity must meet two conditions: the first condition is that the curve does not collide with any obstacle, and the second condition is that the curve does not exceed the boundary of the grid map; if such a curve exists, the path search can be stopped and the final planning path of the Hybrid A Star algorithm can be output.

[0155] Step S401c: perform path planning on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path.

[0156] Optionally, step S401c specifically includes the following steps:

[0157] Step S401c1: construct a continuous curvature target tree according to the second path planning points to form a target tree set.

[0158] In this embodiment, the target tree represents a set of curvature-continuous reverse paths in the target parking space, and the reverse paths are composed of straight segments, clothoid segments, and arc segments. To generate the target tree, a dynamic or kinematic model is needed to describe the dynamic characteristics of the vehicle. Considering the low-speed parking planning scenario, the vehicle kinematic model can be used to fully describe the dynamic characteristics of the vehicle. The vehicle kinematic model is defined as follows:

[0159]

[0160] Among them, (x, y) represents the center position of the vehicle's rear axle, Indicates the vehicle's heading angle; L indicates the vehicle's wheelbase; represents the front wheel steering angle of the vehicle; d is 1 or -1. When d=1, it means the vehicle is moving forward, and when d=-1, it means the vehicle is moving backward; ( )' represents the derivative with respect to the arc length of the path; represents the curvature of the path at (x,y), ; Indicates sharpness, which is the derivative of curvature with respect to the arc length of the path, and the derivative of the front wheel angle with respect to the arc length of the road. related( ); The constraints on curvature and sharpness are , It can be seen that the model not only considers the maximum curvature constraint of the vehicle, but also the maximum curvature change rate constraint (i.e., the front wheel turning angle rate constraint).

[0161] Common curves, such as RS curves, are composed of straight lines and circular arcs. The curvature has a step phenomenon, which does not meet the vehicle kinematics and cannot be used as the vehicle's motion trajectory. A clothoid curve is a curve whose curvature changes linearly along the arc length to meet the vehicle's kinematic constraints. Therefore, in this embodiment, a clothoid curve is used to connect straight lines and circular arcs to generate a node tree.

[0162] The target tree is composed of straight lines, clothoid curves, and circular arcs. Therefore, the difference between each target tree is the difference in the length of the straight line. Loop through the length of the straight line to determine a target tree, and then arrive Loop through with a certain step size to generate different branches, and finally connect arcs to each branch to complete the generation of the target tree.

[0163] like Figure 2 As shown, in this embodiment, the parking of the vehicle can also be considered from the perspective of the parking out of the vehicle (that is, the path planning starting point and the path planning end point can be swapped). The green line in the figure represents the straight part (equivalent to the tree trunk), the blue line represents the clothoid curve (equivalent to the different branches on the tree), and the red line represents the arc. It should be noted that when the horizontal parking space in the figure is out of the garage, it is a series of reverse arcs (red lines in the figure), and the curvature is not 0, so it cannot be directly connected to the target tree. It is necessary to connect a section of the clothoid curve to reduce the curvature to 0 (blue line in the figure) and then connect to the target tree.

[0164] Step S401c2: Establish an optimal target tree cost function based on the target tree set, and select the optimal target tree therein.

[0165] Step S401c3: According to the optimal target tree, establish an optimal node cost function, and select the optimal node (ie, the third path planning point).

[0166] Step S401c4: According to the optimal node, a path between the path planning starting point and the optimal node is planned by a path planning algorithm.

[0167] Step S401d: perform path planning on the third path planning point and the fourth path planning point by using the SpaceInside algorithm to form a third path.

[0168] Optionally, in step S401d, the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking exit direction; parking in is considered from the perspective of parking exit.

[0169] like Figure 3-Figure 7As shown, ABCD represents the parking space, FGG'F represents the boundary abstracted from the real obstacle, bdec represents the four corner points of the vehicle, End represents the parking red point calculated in advance, and OXY represents the parking space coordinate system.

[0170] Optionally, step S401d specifically includes the following steps:

[0171] Step S401d1, the midpoint of line segment CD is set to O, the direction passing through point O and being the same as that of vector OC is the positive direction of Y axis, and the direction passing through point O and being the same as that of vector CA is the positive direction of X axis, and an XOY coordinate system is constructed.

[0172] Step S401d2, initialize the line segment ec to collide with the line segment GG' as the initial position of the vehicle S, wherein the rear axle axis of the vehicle S is O1.

[0173] Step S401d3, the point O1 and the line segment ce have an intersection and are parallel to de as the positive direction of the Y1 axis, and the point O1 and the line segment de have an intersection and are parallel to bd as the positive direction of the X1 axis, and the vehicle coordinate system is constructed.

[0174] Step S401d4: After the steering wheel of vehicle S is rotated in the first direction to the limit position (i.e., the steering wheel is turned to the limit), vehicle S performs a backward operation to determine the steering angle when vehicle point b contacts line segment FF' The steering angle when the vehicle point d contacts the line segment F'G' ,Pick Among them, when the steering wheel of vehicle S rotates to the extreme position in the first direction for the first time, point b contacts the obstacle area before point d, which is optimal, so that the next section of the kneading track can adjust the heading to the maximum extent.

[0175] To ensure that the moving point b of vehicle S first contacts the obstacle area FF'G'G during the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows:

[0176]

[0177] in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the coordinate value of point b on the Y axis when vehicle S reaches the end point of the backward movement for the first time.

[0178] Step S401d5: When the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the backward rotation center of the vehicle S is calculated; wherein, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rotation center of the vehicle S is calculated as follows:

[0179]

[0180] Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinate of vehicle S in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows:

[0181]

[0182] Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows:

[0183]

[0184] in, Indicates the angle of rotation. When the rear end of the vehicle rotates toward the line segment F'G', the rotation angle is positive, negative, rotation angle is negative (in the embodiment, when the vehicle rotates in the counterclockwise direction, the rotation angle is positive, when the vehicle rotates clockwise, the rotation angle is negative).

[0185] Regarding the rotation of the vehicle around the center of rotation, the proof is as follows:

[0186] like Figure 6 As shown, point d rotates around the rotation center T by an angle After reaching point d', point p represents the intersection of the perpendicular line from point T to line segment F'G'. In triangle d'Tp, the length of line segment d'T represents the turning radius of the vehicle corner point d, denoted by R d , the length of line segment pT is recorded as L pT , then the length of line segment d'p is L d'p It can be expressed as:

[0187]

[0188] Among them, L pT It can be expressed as .

[0189] in, , Represent the coordinates of point F' and point T on the X-axis respectively.

[0190] Then, the coordinates of point d' It can be expressed as follows:

[0191]

[0192] Obviously, point d and point d' are on the same circle with point T as the center, as Figure 7 As shown. And by Figure 7 The rotation angle is .

[0193] Step S401d6: According to the , the backward rotation center calculates the backward motion trajectory of the vehicle S and the backward motion end point.

[0194] Step S401d7, set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the steering wheel of the vehicle S in the second direction to the extreme position, and then the vehicle S moves forward to determine the steering angle when the vehicle point d contacts the line segment F'G' The steering angle at which the vehicle point e contacts the line segment GG' ,Pick .

[0195] Step S401d8, when the vehicle contacts the boundary line segment of area FF'G'G, calculate the forward rotation center of vehicle S. Otherwise, take the X-axis coordinate value of point e of vehicle S on the trajectory that is the same as or slightly higher than the X-axis coordinate value of point G, and output the motion trajectory of the SpaceInside algorithm.

[0196] Step S401d9: According to the , the forward rotation center calculates the forward phase motion trajectory of the vehicle S and the forward phase motion end point.

[0197] Step S401d10: Repeat steps S401d4 to S401d9 until the vehicle does not come into contact with the boundary line segment of area FF'G'G, and then output the motion trajectory of the SpaceInside algorithm.

[0198] like Figure 8As shown in the figure, FF'G'G is the abstract obstacle boundary, the blue car frame represents the end pose of parking, the green car frame represents the calculated start kneading pose of SpaceInside algorithm, the black car frame represents the end point of the trajectory when the steering wheel is turned to the right, the purple car frame represents the end point of the trajectory when the steering wheel is turned to the left, Co1 represents the rotation center when retreating, Co2 represents the rotation center when moving forward, and the pink dotted line represents the movement of the corner point e of the vehicle. According to the above, the starting point of SpaceInside is to virtualize the car to the extreme position in front, to ensure that there is no collision, to rotate the maximum angle to ensure no collision, to calculate the position of the rear boundary from the corner point b, and then to move the extreme position to become the starting point of SpaceInside. In the figure, the extreme position represented by the green car frame will collide with the rear boundary first after rotation, so there is no need to move. The green car frame indicates that the initial extreme position of the vehicle is the starting point of SpaceInside.

[0199] Step S401e, perform path planning on the fourth path planning point and the fifth path planning point by the StraightLine algorithm to form a fourth path. After the StraightLine algorithm generates a straight line between the seventh path planning point and the eighth path planning point, the straight line is discretized into a series of path points at equal intervals to obtain the path output by the StraightLine algorithm. The purpose of performing path planning by the StraightLine algorithm is to leave room for adjustment for subsequent control. If the previous tracking error is large, the StraightLine algorithm can be used to correct the position of the vehicle.

[0200] Step S401f: Combine the first path, the second path, the third path and the fourth path to generate a parking planning route.

[0201] Optionally, if the current parking scene type is a non-horizontal double-boundary parking scene, step S4 specifically includes the following steps:

[0202] Step S402a, determine the coordinates of the three path planning points of the current non-horizontal double-boundary parking space scene, and record them as the sixth path planning point, the seventh path planning point and the eighth path planning point respectively; wherein the sixth path planning point is the planning starting point, and the eighth path planning point is the planning end point.

[0203] Step S402b: perform path planning on the sixth path planning point and the seventh path planning point by using the Hybrid A Star algorithm to form a fifth path.

[0204] Step S402c: perform path planning on the seventh path planning point and the eighth path planning point by using the StraightLine algorithm to form a sixth path.

[0205] Step S402d: Combine the fifth path and the sixth path to generate a parking planning route.

[0206] Optionally, in step S4, the planning starting point is set as the parking planning end point, the planning end point is set as the parking planning starting point, the parking planning method remains unchanged, and the parking path planning is performed according to the parking planning starting point, the parking planning end point and the parking planning method to form a parking planning route, and the parking planning route is the parking planning route.

[0207] like Fig. 9 The figure shows a parking planning diagram for a horizontal double-boundary parking space scenario. The black dotted line indicates the key point description, and the red dotted line indicates the planning direction when performing path planning. The parking starting point 1 (i.e., the first path planning point) is planned to the second path planning point 2 by the HybridA Star algorithm to form the first path; the fourth path planning point 4 is planned to the third path planning point 3 by the SpaceInside algorithm through reverse thinking to form the third path; and then the third path planning point 3 is planned to the second path planning point 2 by the TargetTree algorithm to form the second path; the fourth path planning point 4 is planned to the parking end point 5 (i.e., the fifth path planning point) by the StraightLine algorithm to form the fourth path.

[0208] The fourth path mentioned above is to keep a straight line to control tracking and adjust tracking errors, and the third path is to reduce unnecessary parking times. For parking planning in horizontal double-boundary parking spaces, the reason for using four planning algorithms is that in narrow horizontal double-boundary parking spaces, it is difficult for the Hybrid A Star algorithm to obtain path planning results, especially on low-computing platforms equipped with vehicles. At the same time, there may be unnecessary parking times. Therefore, the planning endpoint of the Hybrid A Star algorithm is set outside the parking space, which improves the success rate of parking path planning while minimizing the number of parking times for the vehicle.

[0209] Embodiment 2:

[0210] like Fig.10 As shown, the trajectory planning device of a parking APA multiple mode combination form described in this embodiment specifically includes the following modules:

[0211] The information acquisition module is used to obtain the obstacle information around the vehicle, the parking space information (i.e. the planned end point) and the current position information of the vehicle (i.e. the planned starting point).

[0212] The parking space scene determination module is used to determine the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle.

[0213] Preferably, in the parking space scene determination module, the parking space scene types include: vertical parking spaces, inclined parking spaces and horizontal parking spaces.

[0214] Among them, the vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces.

[0215] Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

[0216] The parking planning method determination module is used to determine the parking planning method according to the parking space scene type.

[0217] Among them, the parking planning methods include two types; among them, the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm, and the second parking planning method includes the search method Hybrid A Star algorithm, the Target Tree algorithm, the SpaceInside algorithm and the StraightLine algorithm.

[0218] Among them, for vertical parking spaces, inclined parking spaces, horizontal line parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning.

[0219] Among them, when planning the parking path through the first parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; at this time, the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm.

[0220] Among them, when planning the parking path through the second parking planning method, the information that needs to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; at this time, the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm.

[0221] The parking route planning module is used to plan a parking path according to the planning starting point, the planning end point and the determined parking planning method to form a parking planning route.

[0222] Optionally, if the current parking space scene type is a horizontal double-boundary parking space scene, the parking route planning module specifically includes the following sub-modules:

[0223] The first submodule of parking route planning is used to determine the coordinates of the five path planning points of the current horizontal double-boundary parking space scene, namely the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point; among which, the first path planning starting point is the planning starting point, and the fifth path planning point is the planning end point.

[0224] The second submodule of parking route planning is used to perform path planning on the first path planning point and the second path planning point through the Hybrid A Star algorithm to form a first path.

[0225] The third submodule of parking route planning is used to perform path planning on the second path planning point and the third path planning point through the TargetTree algorithm to form a second path.

[0226] The fourth submodule of parking route planning is used to perform path planning on the third path planning point and the fourth path planning point through the SpaceInside algorithm to form a third path; wherein the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking exit direction; parking is considered from the perspective of parking exit.

[0227] Among them, ACBD, bdec and FF'G'G are all rectangular areas and the four vertices of each rectangle correspond to each other; among them, ACBD represents the parking space; bdec represents the area where the vehicle is located; FF'G'G represents the boundary abstracted from the obstacle; in the rectangular area FF'G'G, point F and point G are connected through point F' and point G' in turn; in the rectangular area ACBD, point A and point B are connected through point C and point D in turn; the width of the rectangular area FF'G'G is FF' (the length of the line segment FF' is L FF' ), the length is F'G' (the length of line segment F'G' is L F'G' ), the width of the rectangular area ACBD is AC (the length of the line segment AC is L AC ), the length is CD (the length of line segment CD is L CD ), the width of bdec is bd (the length of line segment bd is L bd ), length is de (the length of line segment de is L de ), and L FF' >L AC >L bd , L F'G' >L CD >L de .

[0228] The line segment ec represents the front end of the vehicle, and the line segment bd represents the rear end of the vehicle.

[0229] Among them, the fourth submodule of parking route planning specifically includes the following submodules:

[0230] The fifth module is used to set the midpoint of line segment CD to O, the direction of the line segment passing through point O and being the same as the direction of vector OC is the positive direction of the Y axis, and the direction of the line segment passing through point O and being the same as the direction of vector CA is the positive direction of the X axis, so as to construct an XOY coordinate system.

[0231] The sixth submodule is used to initialize the line segment ec to collide with the line segment GG' as the initial position of the vehicle, wherein the rear axle axis of the vehicle is O1.

[0232] The seventh module is used to construct a vehicle coordinate system with the intersection of point O1 and line segment ce and parallel to de as the positive direction of the Y1 axis, and the intersection of point O1 and line segment de and parallel to bd as the positive direction of the X1 axis.

[0233] The eighth submodule is used to determine the steering angle when the vehicle point b contacts the line segment FF' after the vehicle steering wheel is rotated to the extreme position in the first direction (i.e. the steering wheel is turned to the limit) and the vehicle performs a reverse operation The steering angle when the vehicle point d contacts the line segment F'G' ,Pick Among them, when the steering wheel of the vehicle rotates to the extreme position in the first direction for the first time, point b contacts the obstacle area before point d, which is optimal, so that the next section of the kneading track can adjust the heading to the maximum extent.

[0234] At the same time, in order to ensure that the moving point b of the vehicle first contacts the obstacle area FF'G'G in the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows:

[0235]

[0236] in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the Y-axis coordinate value of point b when the vehicle first reaches the end point of the backward movement.

[0237] The ninth submodule is used to calculate the vehicle's backward rotation center when the vehicle contacts the boundary line segment of the rectangular area FF'G'G; wherein, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the vehicle's rotation center is calculated as follows:

[0238]

[0239] Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinates of the vehicle in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows:

[0240]

[0241] Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows:

[0242]

[0243] in, Indicates the angle of rotation. When the rear end of the vehicle rotates close to line segment F'G', the rotation angle is positive, negative, rotation angle is negative.

[0244] The tenth grandchild module is used according to the , the backward rotation center calculates the backward motion trajectory of the vehicle and the backward motion end point.

[0245] The eleventh module is used to set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the vehicle steering wheel in the second direction to the extreme position, and then the vehicle moves forward to determine the steering angle when the vehicle point d contacts the line segment F'G'. The steering angle at which the vehicle point e contacts the line segment GG' ,Pick .

[0246] The twelfth module is used to calculate the forward rotation center of the vehicle when the vehicle contacts the boundary segment of area FF'G'G. Otherwise, the X-axis coordinate value of the vehicle point e on the trajectory is the same as or slightly higher than the X-axis coordinate value of point G, and the motion trajectory of the SpaceInside algorithm is output.

[0247] The thirteenth grandson module is used according to the , the forward rotation center calculates the vehicle's forward phase motion trajectory and the forward phase motion endpoint.

[0248] The fourteenth grandson module is used to repeatedly execute the fifth grandson module, the sixth grandson module, the seventh grandson module, the eighth grandson module, the ninth grandson module, the tenth grandson module, the eleventh grandson module, the twelfth grandson module and the thirteenth grandson module until the vehicle does not contact the boundary segment of the area FF'G'G, and output the motion trajectory of the SpaceInside algorithm.

[0249] The fifth submodule of parking route planning is used to perform path planning on the fourth path planning point and the fifth path planning point by using the StraightLine algorithm to form a fourth path.

[0250] The sixth submodule of parking route planning is used to combine the first path, the second path, the third path and the fourth path to generate a parking planning route.

[0251] Optionally, if the current parking space scene type is a non-horizontal double-boundary parking space scene, the parking route planning module specifically includes the following sub-modules:

[0252] The seventh submodule of parking route planning is used to determine the coordinates of the three path planning points of the current non-horizontal double-boundary parking space scene, which are respectively recorded as the sixth path planning point, the seventh path planning point and the eighth path planning point; among them, the sixth path planning point is the planning starting point, and the eighth path planning point is the planning end point.

[0253] The eighth submodule of parking route planning is used to perform path planning on the sixth path planning point and the seventh path planning point through the Hybrid A Star algorithm to form a fifth path.

[0254] The ninth submodule of parking route planning is used to perform path planning for the seventh path planning point and the eighth path planning point through the StraightLine algorithm to form a sixth path. After the StraightLine algorithm generates a straight line between the seventh path planning point and the eighth path planning point, the straight line is discretized into a series of path points at equal intervals to obtain the path output by the StraightLine algorithm. The purpose of performing path planning through the StraightLine algorithm is to leave room for adjustment for subsequent control. If the previous tracking error is large, the position of the vehicle can be corrected through the StraightLine algorithm.

[0255] The tenth submodule of parking route planning is used to combine the fifth path and the sixth path to generate a parking planning route.

[0256] The specific embodiments of the present invention are described in detail above, but they are only examples, and the present invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications and substitutions made to the present invention are also within the scope of the present invention. Therefore, the equalization changes and modifications made without departing from the spirit and scope of the present invention should be included in the scope of the present invention.

Claims

1. A method for trajectory planning in the form of a combination of multiple parking APA modes, characterized in that: The specific steps include: Step S1, obtaining information about obstacles around the vehicle, parking space information and the current position information of the vehicle, taking the current position information of the vehicle as the planning starting point, and taking the parking space information as the planning end point; Step S2: determining the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle; Step S3: determining a parking planning method according to the parking space scene type; Step S4, planning a parking path according to the planning starting point, the planning end point and the determined parking planning method to form a parking planning route; In step S3, the parking planning method includes two methods; the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm; the second parking planning method includes the search method Hybrid A Star algorithm, Target Tree algorithm, SpaceInside algorithm and StraightLine algorithm; Among them, for vertical parking spaces, inclined parking spaces, horizontal parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning; When the parking path is planned by the first parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm; When the parking path is planned by the second parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein, the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm; If the current parking space scenario type is a horizontal double-boundary parking space scenario, the parking path planning in step S4 specifically includes the following steps: Step S401a, determining the coordinates of five path planning points of the current horizontal double-boundary parking space scene, which are the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point, wherein the first path planning point is the planning starting point and the fifth path planning point is the planning end point; Step S401b, performing path planning on the first path planning point and the second path planning point by using the Hybrid A Star algorithm to form a first path; Step S401c, performing path planning on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path; Step S401d: performing path planning on the third path planning point and the fourth path planning point by using the SpaceInside algorithm to form a third path; Step S401e, performing path planning on the fourth path planning point and the fifth path planning point by using the StraightLine algorithm to form a fourth path; Step S401f, combining the first path, the second path, the third path and the fourth path in sequence to generate a complete parking planning route; In step S401d, the third path planning point and the fourth path planning point are path planned by using the SpaceInside algorithm to form a third path; wherein the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking exit direction; parking in is considered from the perspective of parking exit; Among them, ACBD, bdec and FF'G'G are all rectangular areas and the four vertices of each rectangle correspond to each other; among them, ACBD represents the parking space; bdec represents the area where the vehicle is located; FF'G'G represents the boundary abstracted from the obstacle; in the rectangular area FF'G'G, point F and point G are connected through point F' and point G' in turn; in the rectangular area ACBD, point A and point B are connected through point C and point D in turn; the width of the rectangular area FF'G'G is FF', and the length of the line segment FF' is L FF' , the length is F'G', the length of line segment F'G' is L F'G' , the width of the rectangular area ACBD is AC, and the length of the line segment AC is L AC , the length is CD, the length of line segment CD is L CD , the width of bdec is bd, and the length of line segment bd is L bd , length is de, the length of line segment de is L de , and L FF' >L AC >L bd , L F'G' >L CD >L de ; Among them, the line segment ec represents the front end of the vehicle, and the line segment bd represents the rear end of the vehicle; Among them, in step S401d, the parking entry is considered from the perspective of the parking exit, that is, the fourth path planning point is used as the planning starting point of the SpaceInside algorithm, and the third path planning point is used as the planning end point of the SpaceInside algorithm to perform path planning, which specifically includes the following steps: Step S401d1, the midpoint of line segment CD is set to O, the positive direction of the Y axis is the direction passing through point O and the same as the direction of vector OC, and the positive direction of the X axis is the direction passing through point O and the same as the direction of vector CA, and an XOY coordinate system is constructed; Step S401d2, initializing line segment ec to collide with line segment GG' as the initial position of the vehicle, wherein the rear axle axis of the vehicle is O1; Step S401d3, the vehicle coordinate system is constructed by passing through the intersection of point O1 and line segment ce and parallel to de as the positive direction of the Y1 axis, and passing through the intersection of point O1 and line segment de and parallel to bd as the positive direction of the X1 axis; Step S401d4: After the vehicle steering wheel is rotated in the first direction to the limit position, the vehicle performs a backward operation to determine the steering angle when the vehicle point b contacts the line segment FF'. The steering angle when the vehicle point d contacts the line segment F'G' ,Pick ; Step S401d5: when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rearward rotation center of the vehicle is calculated; Step S401d6: According to the , the backward rotation center calculates the backward motion trajectory of the vehicle and the backward motion endpoint; Step S401d7, set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the vehicle steering wheel in the second direction to the extreme position, and then move the vehicle forward to determine the steering angle when the vehicle point d contacts the line segment F'G' The steering angle at which the vehicle point e contacts the line segment GG' ,Pick ; Step S401d8, when the vehicle contacts the boundary line segment of area FF'G'G, calculate the forward rotation center of the vehicle, otherwise, take the X-axis coordinate value of the vehicle point e on the trajectory that is the same as or slightly higher than the X-axis coordinate value of point G, and output the motion trajectory of the SpaceInside algorithm; Step S401d9: According to the , the forward rotation center calculates the forward phase motion trajectory of the vehicle and the forward phase motion end point; Step S401d10, repeating steps S401d4 to S401d9 until the vehicle does not come into contact with the boundary line segment of area FF'G'G, and outputting the motion trajectory of the SpaceInside algorithm; The first direction indicates the rotation direction that makes the rear end of the vehicle approach the line segment F'G' during the vehicle's backward movement; the second direction indicates the rotation direction that makes the rear end of the vehicle move away from F'G' during the vehicle's forward movement.

2. The method for trajectory planning of a parking APA multiple mode combination according to claim 1, characterized in that: In step S2, the parking space scene types include: vertical parking space, inclined parking space and horizontal parking space; The vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces; Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

3. The method for trajectory planning of a parking APA combination of multiple modes according to claim 1, characterized in that: If the current parking space scene type is a non-horizontal double-boundary parking space scene, the parking path planning in step S4 specifically includes the following steps: Step S402a, determining the coordinates of three path planning points of the current non-horizontal double-boundary parking space scene, which are the sixth path planning point, the seventh path planning point and the eighth path planning point, wherein the sixth path planning point is the planning starting point and the eighth path planning point is the planning end point; Step S402b, performing path planning on the sixth path planning point and the seventh path planning point by using the Hybrid A Star algorithm to form a fifth path; Step S402c, performing path planning on the seventh path planning point and the eighth path planning point by using the StraightLine algorithm to form a sixth path; Step S402d: Combine the fifth path and the sixth path to generate a complete parking planning route.

4. The method for trajectory planning of a parking APA combination of multiple modes according to claim 1, characterized in that: In step S401c, path planning is performed on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path, which specifically includes the following steps: Step S401c1, constructing a continuous curvature target tree according to the second path planning points to form a target tree set; Step S401c2: establishing an optimal target tree cost function according to the target tree set, and selecting the optimal target tree therein; Step S401c3: Establish an optimal node cost function according to the optimal target tree, and select the optimal node; Step S401c4: According to the optimal node, a path between the path planning starting point and the optimal node is planned by a path planning algorithm.

5. The method for trajectory planning of a parking APA combination of multiple modes according to claim 1, characterized in that: In step S401d4, when the steering wheel of the vehicle rotates to the extreme position in the first direction for the first time, point b contacts the obstacle area before point d, which is optimal, so that the next section of the kneading track can adjust the heading to the maximum extent; In order to ensure that the moving point b of the vehicle first contacts the obstacle area FF'G'G in the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows: in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the Y-axis coordinate value of point b when the vehicle first reaches the end point of the backward movement.

6. The method for trajectory planning of a parking APA combination of multiple modes according to claim 1, characterized in that: In step S401d5, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rotation center of the vehicle is calculated as follows: Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinates of the vehicle in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows: Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows: in, Indicates the angle of rotation. When the rear end of the vehicle rotates toward the line segment F'G', the rotation angle is positive, negative, rotation angle is negative.

7. A device for trajectory planning of parking APA in combination of multiple modes, characterized in that: The modules include: The information acquisition module is used to obtain the obstacle information around the vehicle, the parking space information and the current position information of the vehicle. The current position information of the vehicle is the planning starting point, and the parking space information is the planning end point; A parking space scene determination module, used to determine the parking space scene type of the current parking according to the parking space information and the obstacle information around the vehicle; A parking planning method determination module, used to determine a parking planning method according to the parking space scene type; A parking route planning module, used to plan a parking path according to the planning starting point, the planning end point and the determined parking planning method to form a parking planning route; In the parking planning method determination module, the parking planning methods include two types; wherein the first parking planning method includes the search method Hybrid A Star algorithm and the StraightLine algorithm; the second parking planning method includes the search method Hybrid A Star algorithm, the Target Tree algorithm, the SpaceInside algorithm and the StraightLine algorithm; Among them, for vertical parking spaces, inclined parking spaces, horizontal parking spaces, and horizontal single-boundary parking spaces, the first parking planning method is used for parking planning, and for horizontal double-boundary parking spaces, the second parking planning method is used for parking planning; When the parking path is planned by the first parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein the planning end point of the Habrid A Star algorithm is the planning starting point of the StraightLine algorithm; When the parking path is planned by the second parking planning method, the information to be obtained includes obstacle information, the planning starting point and planning end point of the Habrid A Star algorithm, the planning starting point and planning end point of the TargetTree algorithm, the planning starting point and planning end point of the SpaceInside algorithm, and the planning starting point and planning end point of the StraightLine algorithm; wherein, the planning end point of the Habrid A Star algorithm is the planning starting point of the TargetTree algorithm, the planning end point of the TargetTree algorithm is the planning starting point of the SpaceInside algorithm, and the planning end point of the SpaceInside algorithm is the planning starting point of the StraightLine algorithm; If the current parking space scenario type is a horizontal double-boundary parking space scenario, the parking route planning module specifically includes the following submodules: The first submodule of parking route planning is used to determine the coordinates of five path planning points of the current horizontal double-boundary parking space scene, namely the first path planning point, the second path planning point, the third path planning point, the fourth path planning point and the fifth path planning point, wherein the first path planning starting point is the planning starting point and the fifth path planning point is the planning end point; The second submodule of parking route planning is used to perform path planning on the first path planning point and the second path planning point by using the Hybrid A Star algorithm to form a first path; The third submodule of parking route planning is used to perform path planning on the second path planning point and the third path planning point by using the TargetTree algorithm to form a second path; The fourth submodule of parking route planning is used to perform path planning on the third path planning point and the fourth path planning point by using the SpaceInside algorithm to form a third path; The fifth submodule of parking route planning is used to perform path planning on the fourth path planning point and the fifth path planning point by using the StraightLine algorithm to form a fourth path; A sixth parking route planning submodule, configured to combine the first path, the second path, the third path and the fourth path to generate a complete parking planning route; In the fourth submodule of parking route planning, the direction from the third path planning point to the fourth path planning point is the parking direction, and the direction from the fourth path planning point to the third path planning point is the parking direction; parking is considered from the perspective of parking. Among them, ACBD, bdec and FF'G'G are all rectangular areas and the four vertices of each rectangle correspond to each other; among them, ACBD represents the parking space; bdec represents the area where the vehicle is located; FF'G'G represents the boundary abstracted from the obstacle; in the rectangular area FF'G'G, point F and point G are connected through point F' and point G' in turn; in the rectangular area ACBD, point A and point B are connected through point C and point D in turn; the width of the rectangular area FF'G'G is FF', and the length of the line segment FF' is L FF' , the length is F'G', the length of line segment F'G' is L F'G' , the width of the rectangular area ACBD is AC, and the length of the line segment AC is L AC , the length is CD, the length of line segment CD is L CD , the width of bdec is bd, and the length of line segment bd is L bd , length is de, the length of line segment de is L de , and L FF' >L AC >L bd , L F'G' >L CD >L de ; Among them, the line segment ec represents the front end of the vehicle, and the line segment represents the rear end of the vehicle; The fourth submodule of parking route planning specifically includes the following submodules: The fifth module is used to set the midpoint of line segment CD to O, the line passing through point O and in the same direction as vector OC is the positive direction of the Y axis, and the line passing through point O and in the same direction as vector CA is the positive direction of the X axis, to construct an XOY coordinate system; The sixth submodule is used to initialize the line segment ec to collide with the line segment GG' as the initial position of the vehicle, wherein the rear axle axis of the vehicle is O1; The seventh grandson module is used to construct the vehicle coordinate system by passing through the intersection of point O1 and line segment ce and parallel to de as the positive direction of Y1 axis, passing through the intersection of point O1 and line segment de and parallel to bd as the positive direction of X1 axis; The eighth submodule is used to determine the steering angle when the vehicle point b contacts the line segment FF' after the vehicle steering wheel is rotated to the extreme position in the first direction and the vehicle performs a backward operation The steering angle when the vehicle point d contacts the line segment F'G' ,Pick ; The ninth submodule is used to calculate the vehicle's backward rotation center when the vehicle contacts the boundary line segment of the rectangular area FF'G'G; The tenth grandchild module is used according to the , the backward rotation center calculates the backward motion trajectory of the vehicle and the backward motion endpoint; The eleventh module is used to set the end point of the backward phase movement as the starting point of the forward phase movement, rotate the vehicle steering wheel in the second direction to the extreme position, and then the vehicle moves forward to determine the steering angle when the vehicle point d contacts the line segment F'G'. The steering angle at which the vehicle point e contacts the line segment GG' ,Pick ; The twelfth module is used to calculate the forward rotation center of the vehicle when the vehicle contacts the boundary line segment of the area FF'G'G. Otherwise, the X-axis coordinate value of the vehicle point e on the trajectory is the same as or slightly higher than the X-axis coordinate value of point G, and output the motion trajectory of the SpaceInside algorithm; The thirteenth grandson module is used according to the , the forward rotation center calculates the forward phase motion trajectory of the vehicle and the forward phase motion end point; A fourteenth grandchild module is used to repeatedly execute the fifth grandchild module, the sixth grandchild module, the seventh grandchild module, the eighth grandchild module, the ninth grandchild module, the tenth grandchild module, the eleventh grandchild module, the twelfth grandchild module and the thirteenth grandchild module until the vehicle does not come into contact with the boundary line segment of the area FF'G'G, and output the motion trajectory of the SpaceInside algorithm; The first direction indicates the rotation direction that makes the rear end of the vehicle approach the line segment F'G' during the vehicle's backward movement; the second direction indicates the rotation direction that makes the rear end of the vehicle move away from F'G' during the vehicle's forward movement.

8. The device for trajectory planning of parking APA in combination of multiple modes according to claim 7, characterized in that: In the parking scene determination module, the parking scene types include: vertical parking, inclined parking and horizontal parking; The vertical parking spaces include vertical line parking spaces, vertical single-boundary parking spaces and vertical double-boundary parking spaces; the diagonal parking spaces include diagonal line parking spaces, diagonal single-boundary parking spaces and diagonal double-boundary parking spaces; the horizontal parking spaces include horizontal line parking spaces, horizontal single-boundary parking spaces and horizontal double-boundary parking spaces; Among them, the line parking space refers to the parking space without obstacles on both sides; the single-boundary parking space refers to the parking space with obstacles on one side; and the double-boundary parking space refers to the parking space with obstacles on both sides.

9. The device for trajectory planning of parking APA multiple combination modes according to claim 7, characterized in that: If the current parking space scenario type is a non-horizontal double-boundary parking space scenario, the parking route planning module specifically includes the following submodules: The seventh submodule of parking route planning is used to determine the coordinates of three path planning points of the current non-horizontal double-boundary parking space scene, which are respectively recorded as the sixth path planning point, the seventh path planning point and the eighth path planning point, wherein the sixth path planning point is the planning starting point and the eighth path planning point is the planning end point; The eighth submodule of parking route planning is used to perform path planning on the sixth path planning point and the seventh path planning point by using the Hybrid A Star algorithm to form a fifth path; The ninth submodule of parking route planning is used to perform path planning on the seventh path planning point and the eighth path planning point by using the StraightLine algorithm to form a sixth path; The tenth submodule of parking route planning is used to combine the fifth path and the sixth path to generate a complete parking planning route.

10. The device for trajectory planning of parking APA in combination of multiple modes according to claim 7, characterized in that: The third submodule of parking route planning specifically includes the following submodules: A first grandchild module, configured to construct a continuous curvature target tree according to the second path planning points to form a target tree set; The second grandchild module is used to establish an optimal target tree cost function according to the target tree set, and select the optimal target tree therein; The third grandchild module is used to establish an optimal node cost function according to the optimal target tree and select the optimal node; The fourth submodule is used to plan the path between the path planning starting point and the optimal node through a path planning algorithm according to the optimal node.

11. The device for trajectory planning of parking APA in multiple combination modes according to claim 7, characterized in that: In the eighth grandchild module, when the steering wheel of the vehicle rotates to the extreme position toward the first direction for the first time, point b contacts the obstacle area before point d, which is optimal, so that the next section of the kneading track can adjust the heading to the maximum extent; In order to ensure that the moving point b of the vehicle first contacts the obstacle area FF'G'G in the first backward phase, the planning starting point of the SpaceInside algorithm is expressed as follows: in, Represents the coordinates of point O1; Indicates a new starting point O 1_new Coordinate of Y F Indicates the coordinate value of point F on the Y axis; It represents the Y-axis coordinate value of point b when the vehicle first reaches the end point of the backward movement.

12. The device for trajectory planning of parking APA multiple combination modes according to claim 7, characterized in that: In the ninth module, when the vehicle contacts the boundary line segment of the rectangular area FF'G'G, the rotation center of the vehicle is calculated as follows: Among them, (X O ,Y O ) represents the coordinates of the rotation center; (X rear ,Y rear ) represents the coordinates of the vehicle in the OXY coordinate system; θ represents the heading of the vehicle; R represents the minimum turning radius of the vehicle; k is a coefficient, which is specifically expressed as follows: Among them, by determining the rotation angle and the center of rotation in each forward or backward stage, the motion trajectory of the vehicle is obtained, which is specifically expressed as follows: in, Indicates the angle of rotation. When the rear end of the vehicle rotates toward the line segment F'G', the rotation angle is positive, negative, rotation angle is negative.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a method for trajectory planning in a combination of multiple parking APA modes as described in any one of claims 1 to 6 is implemented.

14. An electronic device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements a method for trajectory planning in the form of a combination of multiple parking APA modes as described in any one of claims 1 to 6.

Images