CN111152784B - A local path planning method for intelligent valet parking - Google Patents

Abstract

The invention discloses a local path planning method for intelligent valet parking. Including inbound and outbound path planning, intersection right-angle turning path planning: According to the calculated minimum lateral distance, turning radius and collision risk, determine the inbound and outbound control points and generate the inbound and outbound driving trajectory; take the inflection point in the known global path as a control point , and then determine the remaining two control points according to the steering radius R′ of the vehicle turning at a right angle, and generate a right-angle turning driving trajectory at the intersection; finally, plan the path to avoid obstacles. The present invention designs a local path planning method that meets the requirements of the scene for the scene of vehicles entering and leaving the warehouse, the scene of straight-line driving, the scene of turning at an intersection and the scene of vehicle obstacle avoidance in the intelligent valet parking, so that the vehicle can successfully arrive at the target parking space or leave smoothly. PARKING LOT.

Description

A local path planning method for intelligent valet parking

technical field

The invention relates to the fields of intelligent valet parking, path planning and the like, in particular to the field of local path planning of intelligent valet parking.

Background technique

Valet parking means that the smart car automatically drives into the parking lot from the drop-off area according to the parking instruction, and finds an empty parking space for parking; when the vehicle sends the recall instruction, the vehicle can automatically return to the passenger area according to the recall instruction. There is no human involvement in the process. Local path planning technology is a dynamic and real-time planning technology that relies on the environmental information provided by sensors and based on the global path. It is the core component of the autonomous system of intelligent vehicles. It is mainly responsible for planning a safe and smooth local path in real time to ensure the safety and stability of unmanned vehicles. Therefore, the quality of local path planning technology directly affects the performance of the entire autonomous system, so the research of local path planning technology is of great significance for the implementation of valet parking system.

At present, there are few local path planning schemes for intelligent valet parking in academia, and the existing partial path planning schemes are not suitable for intelligent valet parking scenarios.

SUMMARY OF THE INVENTION

Aiming at the above-mentioned shortcomings of the prior art, the present invention designs a local path planning scheme for intelligent valet parking, and performs corresponding local path planning schemes for vehicles leaving the warehouse, driving in a straight line, turning at intersections, avoiding obstacles and entering the warehouse. method, so that the vehicle successfully arrives at the target parking space or smoothly leaves the parking lot. The scheme includes: 1-1 access and exit path planning, 1-2 right-angle turning path planning at intersections, and 1-3 obstacle avoidance path planning.

The 1-1 inbound and outbound path planning steps are as follows:

1-1-1 Calculate the minimum lateral distance when vehicles enter and exit the warehouse safely

Among them, R _V is the minimum turning radius of the vehicle itself, W _V is the width of the vehicle, W _P is the width of the parking space, and α′ is the safety distance between the side rear corner and the edge of the parking space during the process of entering and leaving the garage.

1-1-2 Calculate the turning radius R of the vehicle entering and leaving the warehouse:

R ² +[(W _V +2α′)-(2L ^VR +W _P )]R+(L ^VR ) ² +W _P ² /4-(W _V /2+α′) ² =0

Among them, L ^VR is the actual lateral distance when the vehicle enters and exits the warehouse safely.

1-1-3 Determine whether there is a risk of collision between the head of the vehicle and the front side curb of the warehouse:

Among them, L _V is the width of the vehicle, l is the distance from the midpoint of the rear axle of the vehicle to the rear contour line of the vehicle, α″ is the safety distance between the front side angle of the vehicle and the front side curb of the parking space during the process of entering and leaving the garage, and _LR is the road width.

1-1-4 Calculate the coordinates of the inbound and outbound control points and generate the inbound and outbound trajectory:

1) The coordinates of the inbound and outbound control points are A ₁ (W _P /2+R,L _P +L ^VR ), A ₂ (W _P /2,L _P +L ^VR ), A ₃ (W _P /2,L _P -R+L ^VR ), A ₄ (W _P ,α″′+l); where L _P is the length of the parking space, the safety distance between the rear side of the vehicle and the edge of the storage space after the α″′ vehicle is stored;

2) Based on the second-order Bezier curve and the coordinates of the control points A ₁ , A ₂ , and A ₃ , the vehicle entry and exit curve trajectory is generated:

T ₁ (x)=(1-x ² )A ₁ +2x(1-x)A ₂ +x ² A ₃ , x∈[0,1]

3) Based on the first-order Bezier curve and the control points A ₃ , A ₄ , the straight line trajectory of vehicles entering and leaving the warehouse is generated:

T ₂ (x)=(1-x)A ₃ +xA ₄

The steps for planning the right-angle turn path at the 1-2 intersection are as follows:

1-2-1 Take the inflection point in the known global path as a control point B ₁ (a,b), and then determine the remaining two control points B ₂ (aR',b), B according to the steering radius R' of the vehicle turning at a right angle ₂ (a,bR′);

1-2-2 Based on the second-order Bezier curve and the coordinates of the control points B ₁ , B ₂ , and B ₃ , generate a right-angle turn driving trajectory at the intersection:

T ₃ (y)=(1-y ² )B ₁ +2y(1-y)B ₂ +y ² B ₃ , y∈[0,1]

The steps of 1-3 obstacle avoidance path planning are as follows:

1-3-1 Add a length from the midpoint of the rear axle of the vehicle to the front end of the vehicle from the end of the obstacle close to the vehicle, and increase the width of the two sides of the obstacle by half the length of the vehicle body. At the same time, in order to ensure the driving safety, a certain safety distance is added to the front and side of the obstacle, so as to ensure that the final generated path meets the distance requirements between the actual vehicle and the obstacle;

1-3-2 Based on the polygonal path of obstacle avoidance planned by the path search algorithm, select the starting point _C1 of obstacle avoidance, the contact point _C2 of the front end of the obstacle, the turning point C3 at the corner _of the obstacle, and the completion point of obstacle avoidance C ₄ is used as the control point of the third-order Bezier curve to plan the driving path to avoid obstacles:

T ₄ (z)=C ₁ (1-z) ³ +3C ₂ (1-z) ² z+3C ₃ (1-z)z ² +C ₄ z ³ , z∈[0,1]

The present invention has the following technical effects: the present invention respectively designs a local path planning method that meets the requirements of the scene for the scene of vehicles entering and leaving the warehouse, the scene of straight-line driving, the scene of turning at the intersection and the scene of vehicle obstacle avoidance in the intelligent valet parking, so that the vehicle can Successfully arriving at the target parking space or successfully leaving the parking lot, the present invention is especially suitable for intelligent valet parking scenarios.

Description of drawings

Fig. 1 is the method schematic diagram of the present invention;

Fig. 2 is the storage and exit path planning diagram of the present invention;

Fig. 3 is an intersection right-angle turn path diagram of the present invention;

Fig. 4 is the obstacle avoidance path diagram of the present invention;

Detailed ways

In order to have a clearer understanding of the technical features, objects and effects of the present invention, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings;

As shown in FIG. 1 , an embodiment of the present invention provides a local path planning scheme for intelligent valet parking. The method includes three planning schemes: inbound and outbound path planning, intersection right-angle turning path planning, and obstacle avoidance path planning. After the smart car enters the parking lot, it obtains information on available parking spaces and global route information for admission. Based on this information, the smart car analyzes the path information at the next moment, determines the road type and performs corresponding local path planning. When the vehicle reaches the end of the local path, it starts to analyze the next global path, and the above process is repeated until the smart car reaches the target location and completes parking or leaves the parking lot smoothly.

The above three local path planning methods are as follows:

1) Vehicle in and out path planning

和入库转向半径R。然后，判断车辆头部与库位前侧路牙是否有碰撞风险；若无风险，则确定入库控制点A₁、A₂、A₃和A₄，生成车辆入库曲线轨迹T₁(x)和入库直线轨迹T₂(x)。车辆出库过程与入库过程一致。As shown in Figure 2, take the vehicle warehousing as an example. First of all, according to the vehicle's own turning radius R _V , vehicle width W _V , parking space width W _P , the actual lateral distance L ^VR of the vehicle and the safety distance α′ between the side rear corner and the edge of the parking space in the process of entering and leaving the vehicle, Calculate the minimum lateral distance for safe storage of vehicles

and the inbound turning radius R. Then, determine whether there is a collision risk between the head of the vehicle and the front side curb of the storage position; if there is no risk, determine the storage control points A ₁ , A ₂ , A ₃ and A ₄ , and generate the vehicle storage curve trajectory T ₁ (x ) and the incoming straight line trajectory T ₂ (x). The vehicle outbound process is the same as the inbound process.

Specific steps are as follows:

1-1-1 Calculate the minimum lateral distance when vehicles enter and exit the warehouse safely

As shown in Figure 2, it can be known from the Pythagorean theorem

BD ² +CD ² =CB ² (1)

which is:

(RW _P /2) ² +(RL ^VR ) ² =(RW _V /2-α′) ² , R>R _V (2)

即:Among them, L ^VR is the actual lateral distance when the vehicle enters and exits the warehouse safely. From the monotonicity, it can be obtained that L ^VR and R are positively correlated. Therefore, when the minimum turning radius of the vehicle is taken in the above formula, the minimum lateral distance of the vehicle when entering and leaving the warehouse can be calculated.

which is:

1-1-2 Calculate the turning radius R of the vehicle entering and leaving the warehouse:

When it is determined that the lateral distance of the vehicle is greater than the minimum lateral distance, the steering radius of the vehicle can be calculated according to the current lateral distance.

Expand formula (2) to get:

R ² +[(W _V +2α′)-(2L ^VR +W _P )]R+(L ^VR ) ² +W _P ² /4-(W _V /2+α′) ² =0(4)

1-1-3 Determine whether there is a risk of collision between the head of the vehicle and the front side curb of the warehouse:

After the vehicle meets the minimum lateral distance, it is still possible that the front side of the vehicle collides with the opposite curb or obstacle during the parking process. The maximum lateral distance of the vehicle travel area is calculated and compared to the road width.

During the turning process, the farthest distance L _max from the center B to the front side of the vehicle is:

In order to ensure that there is no danger of collision on the front side of the vehicle, it is necessary to meet:

L _max -CD+α″<L _R (6)

From formula (5) and formula (6), we can get:

Among them, L _V is the width of the vehicle, l is the distance from the midpoint of the rear axle of the vehicle to the rear contour line of the vehicle, α″ is the safety distance between the front side angle of the vehicle and the front side curb of the parking space during the process of entering and leaving the garage, and _LR is the road width.

1-1-4 Calculate the coordinates of the inbound and outbound control points and generate the inbound and outbound trajectory:

1) The coordinates of the inbound and outbound control points are A ₁ (W _P /2+R,L _P +h), A ₂ (W _P /2,L _P +h), A ₃ (W _P /2,L _P -R +h), A ₄ (W _P ,α″′+l);

2) Based on the second-order Bezier curve and the coordinates of the control points A ₁ , A ₂ , and A ₃ , the vehicle entry and exit curve trajectory is generated:

T ₁ (x)=(1-x ² )A ₁ +2x(1-x)A ₂ +x ² A ₃ , x∈[0,1]

3) Based on the first-order Bezier curve and the control points A ₃ , A ₄ , the straight line trajectory of vehicles entering and leaving the warehouse is generated:

T ₂ (x)=(1-x)A ₃ +xA ₄

2) Right-angle turn path planning at intersections

As shown in Fig. 3, the inflection point in the known global path is taken as a control point B ₁ , and then the control points B ₂ and B ₂ (that is, the other two parts of the right triangle formed by the right-angle turn) are determined according to the steering radius R' of the vehicle at a right angle. points), and then generate a right-angle turn driving trajectory T ₃ (y) at the intersection.

Specific steps are as follows:

1-2-1 Take the inflection point in the known global path as a control point B ₁ (a,b), and then determine the remaining two control points B ₂ (aR',b), B according to the steering radius R' of the vehicle turning at a right angle ₂ (a,bR′);

1-2-2 Based on the second-order Bezier curve and the coordinates of the control points B ₁ , B ₂ , and B ₃ , generate a right-angle turn driving trajectory at the intersection:

T ₃ (y)=(1-y ² )B ₁ +2y(1-y)B ₂ +y ² B ₃ , y∈[0,1]

3) Path planning to avoid obstacles

4 is the obstacle avoidance path diagram of the present invention, the end of the obstacle close to the vehicle is increased by a length from the midpoint of the rear axle of the vehicle to the front end of the vehicle, and the width of the two sides of the obstacle is increased by half the length of the vehicle body. At the same time, in order to ensure the driving safety, a certain safety distance is added to the front and side of the obstacle, so as to ensure that the final generated path meets the distance requirements between the actual vehicle and the obstacle;

Based on the broken line path of obstacle avoidance planned by the path search algorithm, the starting point _C1 of obstacle avoidance, the contact point _C2 of the front end of the obstacle, the turning point C3 at the corner of the obstacle, and the completion point _C4 of obstacle avoidance are selected as the _third -order The control points of the Bezier curve, planning out the obstacle avoidance path T ₄ (z):

T ₄ (z)=C ₁ (1-z) ³ +3C ₂ (1-z) ² z+3C ₃ (1-z)z ² +C ₄ z ³ , z∈[0,1]

In the description of this specification, reference to the terms "one embodiment," "some embodiments," "exemplary embodiment," "example," "specific example," or "some examples", etc., is meant to incorporate the embodiments A particular feature, structure, material, or characteristic described by an example or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, The scope of the invention is defined by the claims and their equivalents.

Claims (3)

1. a kind of intelligent valet parking local path planning method, is characterized in that, comprises the following steps: 1-1 Warehouse entry and exit path planning: first calculate the minimum lateral distance when vehicles enter and exit the warehouse safely

Then obtain the turning radius R of the vehicle entering and leaving the warehouse, and determine whether there is a risk of collision between the head of the vehicle and the front side curb of the warehouse; finally calculate the coordinates of the entry and exit control points and generate the entry and exit trajectory; 1-2 Right-angle turning path planning at intersection: take the inflection point in the known global path as a control point, and then determine the remaining two control points according to the steering radius R′ of the vehicle turning at a right angle; then generate a right-angle turning driving trajectory at the intersection; 1-3 Avoid obstacle path planning; The specific steps of the 1-1 inbound and outbound path planning are as follows: 1-1-1 Calculate the minimum lateral distance when vehicles enter and exit the warehouse safely

Among them, R _V is the minimum turning radius of the vehicle itself, W _V is the width of the vehicle, W _P is the width of the parking space, and α′ is the safety distance between the rear corner of the vehicle and the edge of the parking space during the process of entering and leaving the garage; 1-1-2 Calculate the turning radius R of the vehicle entering and leaving the warehouse: R ² +[(W _V +2α′)-(2L ^VR +W _P )]R+(L ^VR ) ² +W _P ² /4-(W _V /2+α′) ² =0 Among them, L ^VR is the actual lateral distance when the vehicle enters and exits the warehouse safely; 1-1-3 Determine whether there is a risk of collision between the head of the vehicle and the front side curb of the warehouse:

Among them, L _V is the width of the vehicle, l is the distance from the midpoint of the rear axle of the vehicle to the rear contour line of the vehicle, α″ is the safety distance between the front side angle of the vehicle and the front side curb of the parking space during the process of entering and leaving the garage, and _LR is the road width; 1-1-4 Calculate the coordinates of the inbound and outbound control points and generate the inbound and outbound trajectory: 1) The coordinates of the inbound and outbound control points are A ₁ (W _P /2+R,L _P +L ^VR ), A ₂ (W _P /2,L _P +L ^VR ), A ₃ (W _P /2,L _P -R+L ^VR ), A ₄ (W _P ,α″′+l), where L _P is the length of the parking space, the safety distance between the rear side of the vehicle and the edge of the storage space after the α″′ vehicle is stored; 2) Based on the second-order Bezier curve and the coordinates of the control points A ₁ , A ₂ , and A ₃ , the vehicle entry and exit curve trajectory is generated: T ₁ (x)=(1-x ² )A ₁ +2x(1-x)A ₂ +x ² A ₃ , x∈[0,1] 3) Based on the first-order Bezier curve and the control points A ₃ , A ₄ , the straight line trajectory of vehicles entering and leaving the warehouse is generated: T ₂ (x)=(1-x)A ₃ +xA ₄ . 2. a kind of intelligent valet parking local path planning method according to claim 1, is characterized in that, described 1-2 intersection right-angle turning path planning concrete steps are as follows: 1-2-1 Take the inflection point in the known global path as a control point B ₁ (a,b), and then determine the remaining two control points B ₂ (aR',b), B according to the steering radius R' of the vehicle turning at a right angle ₂ (a,bR′); 1-2-2 Based on the second-order Bezier curve and the coordinates of the control points B ₁ , B ₂ , and B ₃ , generate a right-angle turn driving trajectory at the intersection: T ₃ (y)=(1-y ² )B ₁ +2y(1-y)B ₂ +y ² B ₃ , y∈[0,1]. 3. a kind of intelligent valet parking local path planning method according to claim 1, is characterized in that, described 1-3 avoid obstacle path planning concrete steps are as follows: 1-3-1 Add a length from the midpoint of the rear axle of the vehicle to the front end of the vehicle from the end of the obstacle close to the vehicle, and increase the width of the two sides of the obstacle by half the length of the vehicle body; at the same time, in order to ensure driving safety, in The front end and side of the obstacle are respectively increased by a certain safety distance, so as to ensure that the final generated path meets the distance requirement between the actual vehicle and the obstacle; 1-3-2 Based on the polygonal path of obstacle avoidance planned by the path search algorithm, select the starting point _C1 of obstacle avoidance, the contact point _C2 of the front end of the obstacle, the turning point C3 at the corner _of the obstacle, and the completion point of obstacle avoidance C ₄ is used as the control point of the third-order Bezier curve to plan the driving path to avoid obstacles: T ₄ (z)=C ₁ (1-z) ³ +3C ₂ (1-z) ² z+3C ₃ (1-z)z ² +C ₄ z ³ , z∈[0,1].

Images