Route planning system and method for autonomous vehicles

Abstract

The application discloses a path planning system and method for a self-driving vehicle, belonging to the path planning field, and utilizing at least one sensor to detect surrounding environment information of a host vehicle, and converting the surrounding environment information into an overhead view, which includes coordinate information of each coordinate point. The system discriminates and marks lane edges, lane lines and other vehicles in the overhead view according to the coordinate information, and then calculates a lane center point and finds a preceding vehicle. Next, the system calculates a speed of the preceding vehicle according to a position of the preceding vehicle, and estimates a predicted path of the preceding vehicle. If the predicted path of the preceding vehicle is the same as a driving path of the host vehicle, the preceding vehicle is taken as a path reference point; otherwise, a lane edge is taken as a path reference line to calculate a final path of the host vehicle. The application can plan a path through point cloud data, greatly reduces the cost required for recording high-precision map data, and reduces the storage space occupied by data.

Description

Technical Field

[0001] This invention relates to the field of route planning, and in particular to a route planning system and method for autonomous vehicles. Background Technology

[0002] In recent years, self-driving car technology has gradually matured, and related open-source self-driving car software has entered the market, lowering the barrier to entry for self-driving car development. Currently, most mainstream self-driving car technologies rely on high-precision map data recorded by GPS location or lane line detection to obtain the optimal path.

[0003] The lane detection method for obtaining the optimal path has a significant drawback: not all environments have lane markings, such as intersections and parking lots. Detection will fail in these unmarked areas, thus limiting the effectiveness of lane detection methods due to environmental constraints.

[0004] The method of calculating the optimal route using high-precision map data requires first collecting complete road information using vehicles equipped with stereo cameras to identify road features that aid in positioning. These features include buildings, traffic signs, streetlights, and road markings such as lane lines, directional arrows, and pedestrian crossings. Combining this road information map data with GPS positioning data generates an accurate route image. However, its biggest problem is that it becomes unusable if positioning fails. This is especially true when the vehicle is at an intersection and there are no lane lines to detect, making route planning impossible. Furthermore, obtaining the map data requires significant manpower and financial investment in measurement, and the data volume is substantial, further increasing costs.

[0005] In view of this, the present invention proposes a path planning system and method for autonomous vehicles to address the deficiencies of the aforementioned known technologies and to meet future needs. Summary of the Invention

[0006] The main objective of this invention is to provide a path planning system and method for self-driving vehicles, which can classify surrounding objects based on echo intensity values ​​instead of relying on high-precision map data. This reduces the manpower and cost of recording high-precision map data, while also reducing the space occupied by the data.

[0007] Another objective of this invention is to provide a route planning system and method for self-driving vehicles that does not rely on a navigation system and can still plan a route through physical detection using light-based sensors when navigation fails.

[0008] Another objective of this invention is to provide a path planning system and method for autonomous vehicles, which, in the absence of lane lines on roads or intersections, determines the lane edges through the surrounding environment, finds the lane center point, and then plans a driving path, thereby significantly improving safety.

[0009] To achieve the above objectives, the present invention provides a path planning system for a self-driving vehicle. This system, installed on a vehicle, includes: at least one sensor for detecting the vehicle's surrounding environment; a bird's-eye view generation module connected to the sensor, receiving the surrounding environment information and converting it into a bird's-eye view, which includes coordinate information for each coordinate point; a category detection module connected to the bird's-eye view generation module, identifying and marking lane edges, lane lines, and the preceding vehicle in the bird's-eye view based on the coordinate information; and a lane center calculation module connected to the category detection module, calculating the lane center based on the marked points in the bird's-eye view. The system calculates a lane center point based on the lane edge and lane lines, identifies a preceding vehicle from other vehicles based on the lane center point and the vehicle's position, and calculates the preceding vehicle's speed based on its position; a preceding vehicle prediction module, connected to the category detection module and the lane center calculation module, estimates a predicted path of the preceding vehicle using a vehicle kinematics model; and a path planning module, connected to the preceding vehicle prediction module, uses the preceding vehicle as a path reference point if its predicted path is the same as the vehicle's path, and uses the lane edge as a path reference line to calculate the vehicle's final path if its predicted path is different from the vehicle's path or if there is no preceding vehicle.

[0010] According to an embodiment of the present invention, the sensor is a light-emitting diode (LED). The sensor presents the surrounding environment information of the vehicle as a point cloud map, and the bird's-eye view generation module then uses a rotating axis formula to convert the point cloud map into a bird's-eye view.

[0011] According to embodiments of the present invention, the coordinate information includes the coordinate values ​​of coordinate points, the shape formed by the coordinate points, the density of the number of points, the height of the object formed by the coordinate points, or the echo intensity value of each coordinate point.

[0012] According to an embodiment of the present invention, the echo intensity value is preset with multiple intervals, and the coordinates of the echo intensity values ​​in different intervals are displayed on the bird's-eye view in different colors.

[0013] According to an embodiment of the present invention, the category detection module filters the coordinate information to remove noise, and then identifies the lane edge, lane line and the vehicle in front in the bird's-eye view based on the coordinate information.

[0014] According to an embodiment of the present invention, the category detection module uses a Kalman filter to filter the coordinate information.

[0015] According to an embodiment of the present invention, the lane center calculation module finds a drivable range based on the lane edge and lane lines in the bird's-eye view, and then takes the center point of two adjacent lane lines as the lane center point, or takes the average value of the lane line and the lane edge as the lane center point.

[0016] According to an embodiment of the present invention, after the lane center calculation module obtains the position of the preceding vehicle marked in the bird's-eye view, it calculates the speed of the preceding vehicle based on the position of the preceding vehicle in at least two bird's-eye views over a continuous period of time.

[0017] According to an embodiment of the present invention, the forward vehicle prediction module further establishes a driving behavior interest region by marking lane lines, and then predicts the behavior of the forward vehicle, including going straight or turning, based on the predicted path and the driving behavior interest region.

[0018] According to an embodiment of the present invention, when the predicted path of the preceding vehicle is the same as the driving path of the current vehicle, the path planning module uses the preceding vehicle as a path reference point and combines it with the position of the current vehicle and the center point of the lane line and the speed of the preceding vehicle to calculate the final path of the current vehicle.

[0019] According to an embodiment of the present invention, when the predicted path of the preceding vehicle is different from the driving path of the current vehicle, the path planning module uses the lane edge as a path reference line and calculates an edge curvature based on the path reference line to calculate the final path of the current vehicle.

[0020] The present invention also provides a path planning method for a self-driving vehicle, comprising the following steps: detecting the surrounding environment information of a vehicle using at least one sensor; converting the surrounding environment information into a bird's-eye view, the bird's-eye view including the coordinate information of each coordinate point; identifying and marking the lane edges, lane lines, and other vehicles in the bird's-eye view based on the coordinate information; calculating a lane center point based on the marked lane edges and lane lines in the bird's-eye view; identifying a preceding vehicle from among the other vehicles based on the lane center point and the position of the vehicle; calculating the speed of the preceding vehicle based on the marked position of the preceding vehicle; estimating a predicted path of the preceding vehicle using a vehicle kinematics model; and if the predicted path of the preceding vehicle is the same as the driving path of the vehicle, using the preceding vehicle as a path reference point; if the predicted path of the preceding vehicle is different from the driving path of the vehicle or there is no preceding vehicle, using the lane edge as a path reference line to calculate a final path of the vehicle.

[0021] According to an embodiment of the present invention, the step of calculating the speed of the vehicle ahead based on the position of the vehicle ahead further includes the following step: calculating the speed of the vehicle ahead based on the position of the vehicle ahead in at least two bird's-eye view images over a continuous period of time.

[0022] According to an embodiment of the present invention, the step of estimating a predicted path of a preceding vehicle using a vehicle kinematics model further includes the following steps: establishing a region of interest for driving behavior using marked lane lines, and then predicting the behavior of the preceding vehicle, including going straight or turning, based on the predicted path and the region of interest for driving behavior.

[0023] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: when there are no lane lines on the road or at the intersection, the edge of the lane can be identified by the surrounding environment, and the center point of the lane can be found, thereby planning the driving path, which greatly improves safety. It does not require high-precision map data or GPS. The final path of the vehicle can be calculated by classifying the surrounding environmental objects by the echo intensity value, which greatly reduces the cost required to record high-precision map data, reduces the storage space occupied by the data, and can still be used normally in the basement where there is no GPS. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a block diagram of the self-driving vehicle path planning system of the present invention;

[0026] Figure 2 This is a flowchart of the path planning method for the self-driving car of the present invention;

[0027] Figure 3 A coordinate diagram illustrating the predicted path of the vehicle ahead;

[0028] Figures 4 to 6 A schematic diagram illustrating an embodiment of the invention applied at a T-junction;

[0029] Figures 7 to 9 A schematic diagram illustrating an embodiment of the invention applied at an intersection;

[0030] Figures 10 to 12 This is a schematic diagram illustrating an embodiment of the invention applied in an underground parking lot.

[0031] Symbol explanation:

[0032] 10…Route planning system for autonomous vehicles, 11…Onboard host, 12…Sensors, 13…Bird's-eye view generation module, 14…Category detection module, 15…Lane center calculation module, 16…Forward vehicle prediction module, 17…Route planning module, 20…Lane lines, 22…Lane edges, 24…Other vehicles, 26…This vehicle, 28…Lane center point. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making progressive efforts are within the scope of protection of the present invention.

[0034] It should be understood that, when used in this specification and the appended claims, the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0035] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0036] It should also be further understood that the term "and / or" as used in this specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0037] This invention provides a route planning system and method for autonomous vehicles. Please refer to [link / reference]. Figure 1 This is an architecture diagram of the path planning system for a self-driving vehicle according to the present invention. The path planning system 10 for the self-driving vehicle is installed on a vehicle (not shown in the figure). The path planning system 10 for the self-driving vehicle includes at least one sensor 12, a bird's-eye view generation module 13, a category detection module 14, a lane center calculation module 15, a forward vehicle prediction module 16, and a path planning module 17. The sensor 12 is connected to the bird's-eye view generation module 13, the bird's-eye view generation module 13 is connected to the category detection module 14, the category detection module 14 is connected to the lane center calculation module 15 and the forward vehicle prediction module 16, the lane center calculation module 15 is connected to the forward vehicle prediction module 16, and the forward vehicle prediction module 16 is connected to the path planning module 17. These modules are installed in an on-board host 11, which includes at least one processor (not shown in the figure). These modules can be implemented by one or more processors.

[0038] Sensor 12 is installed on the vehicle to detect information about the vehicle's surrounding environment. In one embodiment, sensor 12 is a lidar sensor that collects point cloud data of the surrounding environment to generate a point cloud map. Bird's-eye view generation module 13 uses a rotation axis formula to convert the point cloud map into a bird's-eye view, which includes coordinate information for each coordinate point, including the coordinate value of each coordinate point, the shape formed by the coordinate points, the density of the points, the height of the objects formed by the coordinate points, or the echo intensity value of each coordinate point, etc., representing the detection category of physical quantities. Because sensor 12 can filter specific information, it can be used to determine whether there is a vehicle ahead.

[0039] In one embodiment, if sensor 12 is a light source, the point cloud echoes it receives will have different echo intensities depending on the material, color, etc. of the object. Therefore, the echo intensity value can be used to distinguish lane edges, lane lines, or vehicles ahead. Specifically, the echo intensity values ​​can be preset into multiple ranges, and the coordinates of different echo intensity value ranges are displayed on the bird's-eye view with different colors. For example, echo intensity values ​​a to b represent a special colored coating. If they meet the characteristics of low height and elongated shape, they are identified as lane lines or lane edges; echo intensity values ​​c to d represent metal. If they meet the characteristics of medium to high height and cube shape, they are identified as vehicles; and echo intensity values ​​e to f represent vegetation or concrete. If they meet the characteristics of medium to high height and irregular shape, they are identified as bushes or sidewalks. This discrimination step is performed by the category detection module 14.

[0040] The category detection module 14 identifies lane lines, lane edges, and all vehicles based on coordinate information and marks them on a bird's-eye view. For example, it draws lane lines and lane edges on the bird's-eye view and selects all vehicles, including the vehicle itself and the vehicle in front / other vehicles. If sensor 12 is a LiDAR, the category detection module 14 first filters the coordinate information using a Kalman filter to remove noise before identifying lane lines, lane edges, and all vehicles.

[0041] The lane center calculation module 15 calculates a lane center point based on the lane edges and lane lines in the marked bird's-eye view. First, a drivable range is identified from the lane edges and lane lines in the bird's-eye view. Then, the center point of two adjacent lane lines within the drivable range is taken as the lane center point, or the average value of the lane line and the lane edge is taken as the lane center point. Multiple lane center points can be connected to form a lane center line. Since it is known which of the marked vehicles is the current vehicle, given the lane center point, it is also possible to determine which of the other vehicles is the preceding vehicle. Furthermore, after obtaining the position of the preceding vehicle, the lane center calculation module 15 can calculate the speed of the preceding vehicle based on the position of the preceding vehicle in at least two consecutive bird's-eye view images. Therefore, the lane center calculation module 15 outputs information such as the drivable range, lane center point, preceding vehicle position, and preceding vehicle speed.

[0042] The preceding vehicle prediction module 16 uses the position of the preceding vehicle captured from the bird's-eye view and a vehicle kinematics model to estimate a predicted path for the preceding vehicle, assuming a constant speed. The preceding vehicle prediction module 16 also establishes a region of interest (ROI) based on marked lane lines, and then predicts the preceding vehicle's behavior t seconds later, including whether it goes straight or turns, based on the predicted path and the ROI. Therefore, the output of the preceding vehicle prediction module 16 is the predicted behavior of the preceding vehicle.

[0043] The path planning module 17 determines whether the predicted path of the preceding vehicle is the same as that of the current vehicle based on the predicted path and behavior of the preceding vehicle. If the predicted path of the preceding vehicle is the same as that of the current vehicle, for example, if both are turning right, then the path of the preceding vehicle is referenced; in other words, the preceding vehicle is used as a path reference point. Combining the positions of the current vehicle, the preceding vehicle, and the center point of the lane, a path equation for the current vehicle's final path can be calculated. If the predicted path of the preceding vehicle is different from that of the current vehicle, then the lane edge closest to the current vehicle is used as a path reference line to calculate the current vehicle's final path.

[0044] Please also refer to Figure 2 This is a flowchart of the path planning method for the self-driving car of the present invention. In step S10, at least one sensor 12 is used to detect the surrounding environment information of the vehicle. In step S12, the bird's-eye view generation module 13 converts the surrounding environment information into a bird's-eye view, which includes the coordinate information of each coordinate point. In step S14, the category detection module 14 identifies and marks the lane edges, lane lines, and other vehicles in the bird's-eye view based on the coordinate information. In step S16, the lane center calculation module 15 calculates a lane center point based on the marked lane edges and lane lines in the bird's-eye view. Then, the lane center calculation module 15 finds a preceding vehicle from among the other vehicles based on the lane center point and the position of the vehicle, and then calculates the speed of the preceding vehicle based on the marked position of the preceding vehicle. In this step, the lane center calculation module 15 first finds a drivable range from the lane edges and lane lines in the bird's-eye view, and then takes the center point of two adjacent lane lines from this drivable range as the lane center point, or takes the average value of the lane line and the lane edge as the lane center point. In step S18, the preceding vehicle prediction module 16 estimates a predicted path for the preceding vehicle using a vehicle kinematics model, and further predicts the preceding vehicle's behavior, including whether it is going straight or turning. Then, as in step S20, the sensor 12 first determines whether there is a preceding vehicle. If there is, in step S22, the path planning module 17 further determines whether the predicted path of the preceding vehicle is the same as the current vehicle's driving path. If they are the same, as described in step S24, the preceding vehicle is used as a path reference point, combined with the positions of the current vehicle and the lane center point, to calculate the current vehicle's final path. Conversely, if the predicted path of the preceding vehicle is different from the current vehicle's driving path, or if step S20 determines that there is no preceding vehicle, as described in step S26, the lane edge is used as a path reference line to calculate the current vehicle's final path.

[0045] In step S12 above, the bird's-eye view generation module 13 uses the rotation axis formula to convert the point cloud map into a bird's-eye view. The rotation axis formula is as follows (1):

[0046]

[0047] Where (x', y', z') are the original coordinates of the point cloud, and (x, y, z) are the coordinates of the transformed bird's-eye view. (cosαi, cosβi, cosγi) is rewritten as (c 1i ,c 2i ,c 3i ), (i=1,2,3). α, β, γ are the angles by which the original coordinate system will be rotated. (h1,h2,h3) is the position of the new origin in the original coordinate system.

[0048] In step S18 above, the preceding vehicle prediction module 16 estimates the predicted path of the preceding vehicle using a vehicle kinematics model, and further predicts the behavior of the preceding vehicle. For details, please refer to [link / reference needed]. Figure 3 This is a coordinate diagram illustrating the predicted path of the vehicle ahead. A represents the position of the front wheels of the self-driving car model; B represents the position of the rear wheels; C represents the position of the center of mass of the self-driving car model; O is the intersection of OA and OB, which is the instantaneous roll center of the vehicle, and line segments OA and OB are perpendicular to the directions of the two tires, respectively. δ r Rear wheel slip angle, δ f For the front wheel deflection angle, L r L is the length from the rear wheel to the center of mass. f Let be the length from the front wheel to the center of gravity. Therefore, the predicted path of the front vehicle can be expressed by the following equation (2):

[0049]

[0050] in, Let ψ be the first coordinate point of the vehicle in front; β be the heading angle; v be the vehicle speed; and r be the wheel angular velocity.

[0051] In step S24 above, the cubic equation k(s) = as 3 +bs 2 +cs+d finds the final path of this vehicle, as shown in equation (3).

[0052] r(s)=[x(s),y(s),θ(s),k(s)]' (3)

[0053] Where s is time, x is the x-axis coordinate point, y is the y-axis coordinate point, θ is the heading angle of the vehicle, k is the curvature of the intersection, and a, b, c, and d are coefficients. In the case of no vehicle in front, the lane curvature can be obtained from the lane edge, and then the following equations (4) to (8) can be substituted into the cubic equation to obtain the final path formula (3) of the vehicle:

[0054] dx / ds=cosθ(s) (4)

[0055] dy / ds=sinθ(s) (5)

[0056] dθ / ds=k(s) (6)

[0057]

[0058]

[0059] Figures 4 to 6 This is a schematic diagram illustrating an embodiment of the invention applied at a T-junction. Taking LiDAR as an example, the point cloud image is first converted using the rotation axis formula. Figure 4 The bird's-eye view shown is represented in grayscale. Next, the surrounding environment is categorized based on echo intensity values ​​to identify lane lines, lane edges, and all vehicles, such as... Figure 5 As shown, the long dashed line represents lane line 20, the short dashed line represents lane edge 22, and the rectangle represents other vehicles 24. The black dot represents the position of your own vehicle 26. Find the lane center point 28 between the two lane lines 20 or between lane line 20 and lane edge 22, as shown... Figure 5 The triangle markings in the diagram. It should be noted that this lane center point 28 is the first center point after passing through the intersection, and as the vehicle 26 moves, the lane center point 28 will also move at each time t. Multiple lane center points 28 can be connected to form a lane center line. Figure 6 In the middle, it determines whether there are vehicles ahead based on the vehicle's width of 26. If there are vehicles ahead, it predicts driving behavior using vehicle kinematics; otherwise, it extracts the curvature of the nearest lane edge 22 for calculation to find the final path for cornering. Figure 6 The light gray arc-shaped arrow in the image indicates the cornering path of car number 26.

[0060] Figures 7 to 9 This is a schematic diagram illustrating an embodiment of the invention applied at an intersection. Taking LiDAR as an example, the point cloud image is first converted using a rotation axis formula. Figure 7 The bird's-eye view shown is represented in grayscale. Next, the surrounding environment is categorized based on echo intensity values ​​to identify lane lines, lane edges, and all vehicles, such as... Figure 8As shown, the short dashed line represents the lane edge 22, and the rectangular box represents another vehicle 26. The black dot represents the position of this vehicle 26. Next, find the lane center point 28 between the two lane edges 22, as shown... Figure 8 The triangle marker in the diagram. Since vehicle 26 can go straight or turn right, the lane center point 28 of both the straight and right turn paths is simultaneously identified, generating... Figure 8 The two triangle markers in the image. (And...) Figure 6 What they have in common is that these two triangular markers are also the center points 28 of the first lane of each of the two paths after passing through the intersection. Figure 9 In the middle, it determines whether there is a vehicle ahead based on the vehicle's width of 26. If there is a vehicle ahead, it predicts the driving behavior of the vehicle ahead using vehicle kinematics; otherwise, it extracts the curvature of the nearest lane edge 22 for calculation to find the final path for cornering, such as... Figure 9 The light gray straight arrows and light gray curved arrows in the image represent the path taken by vehicle 26 through the intersection.

[0061] Figures 10 to 12 This is a schematic diagram illustrating an embodiment of the invention applied in an underground parking lot. Taking LiDAR as an example, the point cloud image is first converted using the rotation axis formula. Figure 10 The bird's-eye view shown is represented in grayscale. Next, the surrounding environment is categorized based on echo intensity values ​​to identify lane lines, lane edges, and all vehicles, such as... Figure 11 As shown, the short dashed line represents lane edge 22, the rectangular box represents other vehicles 24, and the black dot represents the position of your own vehicle 26. Find the lane center point 28 between the two lane edges 22, as shown... Figure 11 The triangle marker in the image. Figure 12 In the middle, the system determines whether there are vehicles ahead based on the width of the vehicle 26. If there is a vehicle ahead, it predicts the driving behavior of the vehicle ahead using vehicle kinematics; otherwise, it extracts the curvature of the nearest lane edge 22 for calculation to find the final path of the vehicle 26. Figure 12 The light gray straight arrow in the image represents the path of vehicle 26.

[0062] In summary, this invention provides a path planning system and method for self-driving vehicles. It converts point cloud maps obtained from LiDAR into a bird's-eye view using a conversion formula, identifies object categories in the surrounding environment, locates lane lines and lane edges, and calculates the drivable range. Simultaneously, it uses the lane lines of other lanes (i.e., the lane after which the vehicle will go straight or turn) to find the lane center point as the endpoint. If there is a vehicle ahead at the lane center point and it is predicted that the preceding vehicle will follow the same path, then the predicted path of the preceding vehicle is used as the path reference point for this vehicle; otherwise, the lane edges of the environment are referenced to calculate the final path of this vehicle. In this way, the vehicle does not require high-precision map data or GPS; the final path can be calculated using the echo intensity values ​​of the LiDAR point cloud data, significantly reducing the cost of recording high-precision map data, reducing the storage space occupied by the data, and allowing the system to function normally even in a basement without GPS.

[0063] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of the invention. Therefore, all equivalent variations or modifications made in accordance with the features and spirit described in the claims of this invention should be included within the scope of the patent application of this invention.

Claims

1. A route planning system for a self-driving car, installed on a vehicle, characterized in that, The self-driving car's route planning system includes: At least one sensor is used to detect information about the vehicle's surrounding environment; A bird's-eye view generation module is connected to the at least one sensor, receives the surrounding environment information and converts it into a bird's-eye view, which includes the coordinate information of each coordinate point. A category detection module, connected to the bird's-eye view generation module, identifies and marks the lane edges, lane lines, and other vehicles in the bird's-eye view based on the coordinate information of each coordinate point; A lane center calculation module, connected to the category detection module, finds a drivable range based on the lane edges and lane lines marked in the bird's-eye view, calculates a lane center point, finds a preceding vehicle from other vehicles based on the lane center point and the position of the vehicle itself, and calculates the speed of the preceding vehicle based on the position of the preceding vehicle. A preceding vehicle prediction module is connected to the category detection module and the lane center calculation module, and estimates a predicted path of the preceding vehicle through a vehicle kinematics model. A path planning module is connected to the preceding vehicle prediction module. If the predicted path of the preceding vehicle is the same as the driving path of the current vehicle, the preceding vehicle is used as a path reference point. Combined with the positions of the current vehicle and the lane center point, a final path of the current vehicle is calculated. If the predicted path of the preceding vehicle is different from the driving path of the current vehicle or there is no preceding vehicle, the lane edge is used as a path reference line. The edge curvature of the lane edge is calculated using the path reference line. The path reference line is moved to the lane center line and scaled so that the curvature of the path reference line is the same as that of the lane edge and conforms to lane centering, so as to calculate the final path of the current vehicle.

2. The self-driving vehicle route planning system as described in claim 1, characterized in that, At least one of the sensors is LiDAR.

3. The self-driving vehicle path planning system as described in claim 2, characterized in that, The at least one sensor presents the vehicle's surrounding environment information as a point cloud map, and the bird's-eye view generation module then uses a rotation formula to convert the point cloud map into the bird's-eye view.

4. The self-driving vehicle path planning system as described in claim 1, characterized in that, The coordinate information includes the coordinate values ​​of the coordinate points, the shape formed by the coordinate points, the density of the points, the height of the object formed by the coordinate points, or the echo intensity value of each coordinate point.

5. The self-driving vehicle path planning system as described in claim 4, characterized in that, The echo intensity value is preset with multiple intervals, and the coordinates of the echo intensity values ​​in different intervals are displayed in different colors on the bird's-eye view.

6. The self-driving vehicle path planning system as described in claim 4, characterized in that, This category detection module filters the coordinate information to remove noise, and then uses the coordinate information to identify the lane edges, lane lines, and other vehicles in the bird's-eye view.

7. The self-driving vehicle path planning system as described in claim 6, characterized in that, This category detection module uses a Kalman filter to filter the coordinate information.

8. The self-driving vehicle route planning system as described in claim 1, characterized in that, The lane center calculation module finds a drivable range based on the lane edge and lane lines in the bird's-eye view, and then takes the center point of two adjacent lane lines in the drivable range as the lane center point, or takes the average value of the lane line and the lane edge as the lane center point.

9. The self-driving vehicle path planning system as described in claim 1, characterized in that, After obtaining the position of the vehicle in front, the lane center calculation module calculates the speed of the vehicle in front based on the position of the vehicle in front in at least two bird's-eye view images over a continuous period of time.

10. The self-driving vehicle path planning system as described in claim 9, characterized in that, The forward vehicle prediction module also establishes a driving behavior interest area by marking the lane line, and then predicts the behavior of the vehicle in front, including going straight or turning, based on the predicted path and the driving behavior interest area.

11. A route planning method for autonomous vehicles, characterized in that, The route planning method for this self-driving car includes the following steps: Use at least one sensor to detect information about the vehicle's surrounding environment; The surrounding environment information is converted into a bird's-eye view, which includes the coordinate information of each coordinate point. Based on the coordinate information of each coordinate point, identify and mark the lane edges, lane lines, and other vehicles in the bird's-eye view; Calculate a lane center point based on the lane edges and lane lines marked in the bird's-eye view. Based on the lane center point and the position of the vehicle, find the preceding vehicle among other vehicles, and then calculate the speed of the preceding vehicle based on its position. A predicted path for the preceding vehicle is estimated using a vehicle kinematics model; If the predicted path of the preceding vehicle is the same as the driving path of this vehicle, the preceding vehicle is used as a path reference point. Combined with the positions of this vehicle and the center point of the lane, a final path of this vehicle is calculated. If the predicted path of the preceding vehicle is different from the driving path of this vehicle or there is no preceding vehicle, the edge of the lane is used as a path reference line. The edge curvature of the lane edge is calculated using the path reference line. The path reference line is moved to the center line of the lane and scaled so that the curvature of the path reference line is the same as that of the lane edge and conforms to lane centering, so as to calculate the final path of this vehicle.

12. The path planning method for autonomous vehicles as described in claim 11, characterized in that, The vehicle's surrounding environment information is presented as a point cloud map, and a rotation formula is used to convert the point cloud map into this bird's-eye view.

13. The path planning method for autonomous vehicles as described in claim 11, characterized in that, The coordinate information includes the coordinate values ​​of the coordinate points, the shape formed by the coordinate points, the density of the points, the height of the object formed by the coordinate points, or the echo intensity value of each coordinate point.

14. The path planning method for autonomous vehicles as described in claim 13, characterized in that, The echo intensity value is preset with multiple intervals, and the coordinates of the echo intensity values ​​in different intervals are displayed in different colors on the bird's-eye view.

15. The path planning method for autonomous vehicles as described in claim 13, characterized in that, After the coordinate information is filtered to remove noise, the lane edges, lane lines, and other vehicles in the bird's-eye view are identified based on the coordinate information.

16. The path planning method for a self-driving vehicle as described in claim 15, characterized in that, The coordinate information is filtered using a Kalman filter.

17. The path planning method for autonomous vehicles as described in claim 11, characterized in that, The calculation of the lane center point involves the following steps: Identify a drivable area based on the lane edges and lane lines in the bird's-eye view; Within the drivable range, the center point of the lane is taken as the center point of the two adjacent lane lines, or the average value of the lane line and the lane edge is taken as the center point of the lane.

18. The path planning method for autonomous vehicles as described in claim 11, characterized in that, The step of calculating the speed of the vehicle in front based on its position also includes the following steps: Calculate the speed of the vehicle in front based on the position of the vehicle in front in at least two bird's-eye view images over a continuous period of time.

19. The path planning method for autonomous vehicles as described in claim 11, characterized in that, The step of estimating a predicted path for the preceding vehicle using the vehicle's kinematics model also includes the following steps: A driving behavior interest region is established by marking the lane lines, and then the behavior of the vehicle in front is predicted based on the predicted path and the driving behavior interest region, including going straight or turning.

Images