Ramp passage auxiliary driving method

Abstract

The invention discloses a ramp passage auxiliary driving method, which comprises the following steps of: when detecting that a ramp exists within a first distance in front of a current position of a vehicle in a global path, acquiring related information of the ramp; according to the global path plan, determining a local path plan passing through the ramp; selecting a ramp starting point for entering the ramp; detecting whether the vehicle runs on the rightmost slow lane at the current position, and if so, keeping the vehicle running on the lane; if not, changing the lane to the slow lane on the rightmost side and continuously driving; calculating to obtain a deceleration distance, and decelerating the vehicle when the distance between the vehicle and the ramp starting point is the deceleration distance; and when the distance between the vehicle and the starting point of the ramp is not greater than a first distance threshold value, switching the current sensing fusion strategy into a fusion strategy mainly based on the millimeter-wave radar, and controlling the vehicle to enter the ramp and drive out. According to the method, false identification of the auxiliary driving system on the ramp lane line and the road side line can be reduced, the accident risk is reduced, and the success rate of ramp passing and the user experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of assisted driving, and in particular to a ramp passage assisted driving method. Background Art

[0002] In the field of ramp assisted driving, the detection and identification of ramp entrances and exits are currently mainly dependent on sensors such as cameras and radars. These sensors can provide visual and distance information of the vehicle's surroundings, but their performance may be affected by factors such as weather, lighting conditions, and obstacle obstruction.

[0003] Existing assisted driving technologies are deficient in the recognition of ramps, especially Y-shaped ramps and safety islands. They are prone to misrecognition due to the divergence and convergence structure of lane lines, which can cause vehicles to drive along the wrong lane line or even collide. In addition, on some ramps with smaller curve radii, the sensor's recognition of lane lines and road sidelines often has problems. Either the recognition length is insufficient to support the vehicle to pass quickly within a certain speed range; or the lane line or road sideline cannot be recognized at all, causing the assisted driving system to be downgraded or directly exit.

[0004] At present, there has been no detailed research on the methods, processes and control strategies for vehicles to pass through ramps. At the same time, the assisted driving system is not perfect in ramp passage, especially the misidentification problem of Y-shaped ramps, safety islands and other structures. There is a lack of effective control logic optimization and improvement strategies, which limits the success rate of ramp passage and user experience. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a ramp passage assisted driving method in response to the deficiencies of the prior art. The method can reduce the assisted driving system's misrecognition of ramp lane lines and road side lines, reduce the risk of accidents, and improve the success rate of ramp passage and user experience.

[0006] To achieve the above object, according to one aspect of the present invention, a ramp driving assistance method is provided, comprising:

[0007] When a ramp is detected within a first distance ahead of the current position of the vehicle in the global path, obtaining relevant information of the ramp;

[0008] Determine a local path plan for passing through the ramp according to the global path plan and the direction in which the vehicle passes through the ramp; select a ramp starting point for entering the ramp according to the local path plan;

[0009] Detect whether the vehicle is currently in the rightmost slow lane. If so, keep driving in this lane; if not, change lanes to the rightmost slow lane and continue driving;

[0010] The deceleration distance is calculated according to the current speed of the vehicle, the preset deceleration value of the vehicle and the speed limit of the ramp, and the vehicle is decelerated when the distance between the vehicle and the starting point of the ramp is the deceleration distance;

[0011] When the distance between the vehicle and the starting point of the ramp is not greater than a first distance threshold, the current perception fusion strategy is switched to a fusion strategy based on millimeter-wave radar, and a road passing through the ramp is output. According to the local path planning, the vehicle is controlled to enter the ramp and exit along the output road passing through the ramp.

[0012] In the above scheme, the method of outputting the road passing through the ramp by the millimeter-wave radar-based fusion strategy includes:

[0013] Data alignment: Align the processed millimeter-wave radar road edges with the road edges of the high-precision map and the visual road edges to ensure that they are in the same coordinate system;

[0014] Feature matching: performing feature matching on the road sidelines of the three data sources, and inspecting and correcting the detailed attributes of the processed millimeter-wave radar road sidelines;

[0015] Road sideline polynomial fitting: performing cubic polynomial fitting on the corrected and processed millimeter-wave radar road sideline and outputting the fused road sideline;

[0016] Virtualizing the lane line on the other side: based on the fused road edge line, virtualizing the lane line on the other side that is equal in width and parallel;

[0017] Lane line detection output: compare and verify the virtual lane line on the other side with the road edge line of the high-precision map and the visual road edge line to ensure that the distance between the virtual lane line on the other side and the fused road edge line is smaller than the distance between the road edge line of the high-precision map and the visual road edge line and the fused road edge line, and output the virtual lane line on the other side.

[0018] In the above scheme, the processed millimeter-wave radar road edge line is obtained by identifying and processing the road edge line on the side of the direction in which the vehicle passes through the ramp based on the millimeter-wave radar, and the processing method includes:

[0019] Data preprocessing: Obtain millimeter-wave radar point cloud data, use filtering algorithms to remove noise points, and separate ground points from non-ground points in the point cloud through cluster analysis; non-ground points include obstacles and road signs.

[0020] Feature extraction: identifying edge points in the point cloud that may belong to the road edge line, and extracting possible straight line segments from the edge points by linear regression method;

[0021] Road sideline recognition: clustering the extracted straight line segments according to their direction, position and length features to distinguish the left and right sidelines, and confirming the road sideline on the direction in which the vehicle passes the ramp;

[0022] Road model fitting: fitting the clustered straight line segments into a road edge model and verifying the accuracy of the road edge model;

[0023] Post-processing: Smoothing the fitted road edge lines to eliminate small fluctuations and noise; using interpolation to fill in missing edge line data due to occlusion.

[0024] In the above solution, the method for the vehicle to travel on the ramp is:

[0025] In the local path planning through the ramp, a driving trajectory point set of the local path planning, as well as the lateral and longitudinal driving speeds and directions of the vehicle when the vehicle is located at the driving trajectory point set are planned at intervals of 1 m; wherein the driving trajectory point set is located on the center line of the fused road edge line and the virtual lane line on the other side.

[0026] When the vehicle is traveling on the local path planning through the ramp, it is planned that the vehicle always drives along the left / right lane of the road, does not change lanes, and does not overtake during the passage.

[0027] In the above scheme, the width of the virtual equal-width and parallel lane lines on the other side is 3-3.2 meters.

[0028] In the above scheme, the method for calculating the deceleration distance is:

[0029]

[0030] Wherein, d is the deceleration distance; Δt is the time required for the vehicle to decelerate from the current speed to the speed limit of the ramp; v 车 is the current speed of the vehicle; v L is the speed limit of the ramp; a is the preset deceleration value of the vehicle, wherein the preset deceleration value ranges from 0.5g to 1g.

[0031] In the above scheme, the specific method for detecting whether there is a ramp within the first distance in front of the current position of the vehicle is:

[0032] During driving, the vehicle detects whether there is a ramp within the first distance ahead of the current position of the vehicle in the global path at a preset frequency based on the high-precision map module in the intelligent driving domain controller;

[0033] If the returned information is negative, the ramp detection loop continues; if the returned information is positive, the ramp related information is obtained.

[0034] In the above scheme, the global path is the path planned by navigation-assisted driving between any two points within the range of the high-precision map.

[0035] In the above scheme, the relevant information of the ramp is the specific attribute information of the ramp on the high-precision map, including the geographical coordinates of the starting and ending points of the ramp, the ramp category, speed limit information, the geographical coordinates of the path guidance points, and the heading angle information; wherein the ramp category is divided into single driving direction and dual driving direction.

[0036] In the above solution, the method for obtaining the direction of the vehicle passing through the ramp is:

[0037] If the ramp type is single-direction, the direction of the vehicle passing through the ramp is to the right;

[0038] If the ramp category is dual-direction driving, when a vehicle passes through the ramp from the right branch of the ramp, the vehicle passes through the ramp in the right direction; when a vehicle passes through the ramp from the left branch of the ramp, the vehicle passes through the ramp in the left direction.

[0039] In the above solution, the local planning is that the direction of the vehicle passing through the ramp is left or right.

[0040] In the above solution, the first distance threshold is 4m.

[0041] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects compared with the prior art:

[0042] The present invention provides a ramp passage assisted driving method, which obtains ramp-related information based on prior information of a high-precision map to prepare for passage, thereby improving the ramp passage success rate and user experience; prepares for entering a slow lane and decelerating before entering a ramp in advance, thereby improving the ramp passage success rate and user experience; by switching to a fusion strategy based on millimeter-wave radar, the road sideline recognition accuracy is improved, the misidentification of lane lines and road sidelines in the ramp is reduced, the accident risk is reduced, and the ramp passage success rate is improved; before entering the ramp, the deceleration module is used to calculate and reserve sufficient distance and time to start deceleration, so that the user will not feel uncomfortable; in the ramp, by driving along the sideline on the driving direction, there is no need to identify safety islands or change lanes during driving, thereby reducing the risk of collision. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Also, throughout the accompanying drawings, the same reference figures are used to represent the same components. In the drawings:

[0044] Figure 1 The figure is a flow chart of a ramp passage assisted driving method in an embodiment of the present invention.

[0045] Figure 2 Schematic diagram of the starting point of a ramp in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0047] It should be understood that the size of the serial numbers of the steps in the embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0048] The present invention is mainly based on an assisted driving system with high-precision maps and navigation functions, and the sensors carried on the vehicle body include forward-looking and surrounding cameras, forward millimeter-wave radar, millimeter-wave corner radar, high-precision maps, combined inertial navigation equipment, and laser radar (optional) that can be configured as needed. The assisted driving system can use perception fusion strategies to navigate the vehicle.

[0049] Example 1

[0050] The present application provides a ramp driving assistance method. Figure 1 ,include:

[0051] S1, when it is detected in the global path that there is a ramp within a first distance ahead of the current position of the vehicle, obtain relevant information of the ramp.

[0052] In the embodiment of the present application, it can be understood that the global path in this embodiment is the path planned by the navigation-assisted driving between any two points within the range of the high-precision map. The vehicle implemented in this embodiment can complete functions such as automatic cruising on the main road from the starting point A to the end point B, autonomous on- and off-ramps, and entering and exiting the main road.

[0053] Specifically, the specific method for detecting whether there is a ramp within a first distance in front of the current position of the vehicle is:

[0054] During driving, the vehicle detects whether there is a ramp within the first distance ahead of the vehicle's current position in the global path at a preset frequency based on the high-precision map module in the intelligent driving domain controller;

[0055] If the returned information is negative, the ramp detection loop continues; if the returned information is positive, the ramp related information is obtained.

[0056] Specifically, in the embodiment of the present application, the first distance is 2 kilometers, and the preset frequency is 20 Hz; Figure 2 The relevant information of the ramp is the specific attribute information of the ramp on the high-precision map, which can be understood as including but not limited to the starting point a of the ramp. 0右 、a 0左 and end point a e右 、a e左 Geographic coordinates, ramp type, speed limit information v L , geographic coordinates of the path guidance points, heading angles and other information; the ramp categories are divided into single-direction C s and dual travel directions C D .

[0057] S2: confirm the local path planning through the ramp according to the global path planning and the direction of the vehicle passing through the ramp; and select the starting point of the ramp entering the ramp according to the local path planning.

[0058] Specifically, in the embodiment of the present application, the method for obtaining the direction of the vehicle passing through the ramp is:

[0059] If the ramp type is one-way driving direction C s , then the direction of the vehicle passing through the ramp is to the right;

[0060] If the ramp type is dual direction C D When a vehicle passes through the ramp from the right branch of the ramp, the direction of the vehicle passing through the ramp is to the right; when a vehicle passes through the ramp from the left branch of the ramp, the direction of the vehicle passing through the ramp is to the left.

[0061] S3, detecting whether the current position of the vehicle is traveling in the rightmost slow lane, if so, maintaining the lane and continuing to travel; if not, changing lanes to the rightmost slow lane and continuing to travel.

[0062] It is understandable that in the embodiment of the present application, the vehicle needs to detect the lane it is currently in before entering the ramp, and cannot directly turn to enter the ramp. When it is detected that the vehicle is currently in the rightmost slow lane, it continues to drive along the lane; when it is detected that the vehicle is currently in the rightmost slow lane, the autonomous lane change command will be triggered until the vehicle autonomously changes lanes to the rightmost slow lane and continues to drive. This is a preparation for the vehicle to enter the ramp.

[0063] S4, calculating a deceleration distance according to the current speed of the vehicle, a preset deceleration value of the vehicle and the speed limit of the ramp, and decelerating the vehicle when the distance between the vehicle and the starting point of the ramp is the deceleration distance.

[0064] In the embodiment of the present application, the specific method for calculating the deceleration distance is:

[0065]

[0066] Where d is the deceleration distance; Δt is the time required for the vehicle to decelerate from the current speed to the speed limit of the ramp; v 车 is the current speed of the vehicle; v L is the speed limit of the ramp; a is the preset deceleration value of the vehicle. To ensure comfort, the preset deceleration value range is between 0.5g and 1g. Specifically, in the embodiment of the present application, a is 0.75g.

[0067] S5: When the distance between the vehicle and the starting point of the ramp is not greater than the first distance threshold, the current perception fusion strategy is switched to a fusion strategy based on millimeter-wave radar, and the road passing through the ramp is output. According to the local path planning, the vehicle is controlled to enter the ramp and exit along the output road passing through the ramp.

[0068] Specifically, in the embodiment of the present application, the first distance threshold is 4m.

[0069] In an embodiment of the present application, a method for outputting a road passing through a ramp using a fusion strategy based on a millimeter-wave radar includes:

[0070] Data alignment: Align the processed millimeter-wave radar road edges with the road edges of the high-precision map and the visual road edges to ensure that they are in the same coordinate system;

[0071] Feature matching: feature matching of road edges from three data sources, and inspection and correction of detailed attributes of processed millimeter-wave radar road edges;

[0072] Road edge polynomial fitting: Perform cubic polynomial fitting on the corrected processed millimeter wave radar road edge and output the fused road edge;

[0073] The cubic polynomial fit is expressed as: y = ax 3 +bx 2 +cx+d

[0074] Among them, y is the vertical position of the lane line, x is the horizontal position of the vehicle, and a, b, c, d are the coefficients of the polynomial.

[0075] Virtual lane line on the other side: Based on the fused road edge line, virtual lane lines of equal width and parallel to the other side are created:

[0076] y v =y+W=v a x 3 +v b x 2 +v c x+v d

[0077] Among them, y v is the position of the virtual lane line on the other side, W is the width of the road. Since the virtual lane line on the other side is parallel to the original road edge line, their normal vectors are the same, and the coefficients of the cubic polynomial of the virtual lane line can be solved by the coefficients of the cubic polynomial of the original road edge line.

[0078] Lane detection output: Compare and verify the virtual lane line on the other side with the road edge line on the high-precision map and the visual road edge line to ensure that the distance between the virtual lane line on the other side and the fused road edge line is smaller than the distance between the road edge line on the high-precision map and the visual road edge line and the fused road edge line, and output the virtual lane line on the other side.

[0079] Specifically, in the embodiment of the present application, the processed millimeter-wave radar road edge line is obtained by identifying and processing the road edge line on the side of the direction in which the vehicle passes through the ramp based on the millimeter-wave radar. The specific processing method includes:

[0080] Data preprocessing: Obtain millimeter-wave radar point cloud data, use filtering algorithms to remove noise points, and separate ground points from non-ground points in the point cloud through cluster analysis;

[0081] Feature extraction: Identify edge points in the point cloud that may belong to the road edge, and extract possible straight line segments from the edge points using linear regression methods;

[0082] Road edge recognition: Cluster the extracted straight line segments according to their direction, position and length characteristics to distinguish the left and right edges and confirm the road edge on the side where the vehicle passes the ramp;

[0083] Road model fitting: fit the clustered straight line segments into a road edge model and verify the accuracy of the road edge model;

[0084] Post-processing: Smooth the fitted road edges to eliminate small fluctuations and noise; use interpolation to fill in missing edge data caused by occlusion.

[0085] Specifically, in this embodiment, the method for a vehicle to travel on a ramp is:

[0086] In the local path planning through the ramp, the driving trajectory point set of the local path planning and the lateral and longitudinal driving speed and direction of the vehicle when it is located in the driving trajectory point set are planned at intervals of 1m; the driving trajectory point set is located on the fused road edge line and the center line of the virtual lane line on the other side. Specifically, when the vehicle is driving on the local path planning through the ramp, the planned vehicle always drives along the left / right lane of the road, does not change lanes, and does not overtake during the passage.

[0087] It is understandable that, because the general lane width is 3.5-3.75 meters, the width of the virtual equal-width and parallel lane lines on the other side is set to 3-3.2 meters.

[0088] A specific application example of the present application is: a vehicle is driving normally according to a global path planned based on a high-precision map in an intelligent driving domain controller. Now it needs to exit from a ramp. Based on the high-precision map module, a 20Hz frequency is used to detect whether there is a ramp 2 kilometers ahead in the global path. If a ramp with a single driving direction is detected, relevant information of the ramp (ramp starting point a 0右 、a 0左 and end point a e右 、a e左 The geographic coordinates of the ramp are C, which is a single-direction ramp. S 、Speed ​​limit informationv L , geographic coordinates of the path guidance point, heading angle, etc.), determine that the direction of the vehicle passing through the ramp is to the right, and select the starting point of the ramp on the same side as the entry point. Detect that the vehicle is currently driving in the rightmost slow lane and continue to drive along this lane. The current vehicle speed is 120km / h, and the ramp speed limit is v L =40km / h, the vehicle's deceleration preset value is 0.75g, and the calculated deceleration distance is 67.12m. The vehicle is controlled to decelerate when the distance between the vehicle and the starting point of the ramp is 67.12m. When the distance between the vehicle and the starting point of the ramp is no more than 4m, the current perception fusion strategy is switched to a fusion strategy based on millimeter-wave radar, and the road through the ramp is output. The vehicle is controlled to keep to the right and enter the ramp and exit along the output road through the ramp.

[0089] It should be pointed out that, according to the needs of implementation, the various steps described in this application can be divided into more steps, or two or more steps or partial operations of the steps can be combined into a new step to achieve the purpose of the present invention.

[0090] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A ramp driving assistance method, characterized in that: include: When a ramp is detected within a first distance ahead of the current position of the vehicle in the global path, obtaining relevant information of the ramp; Determine a local path plan for passing through the ramp according to the global path plan and the direction in which the vehicle passes through the ramp; select a ramp starting point for entering the ramp according to the local path plan; Check whether the vehicle is currently in the rightmost slow lane. If so, keep driving in this lane; If not, change lane to the rightmost slow lane and continue driving; The deceleration distance is calculated according to the current speed of the vehicle, the preset deceleration value of the vehicle and the speed limit of the ramp, and the vehicle is decelerated when the distance between the vehicle and the starting point of the ramp is the deceleration distance; When the distance between the vehicle and the starting point of the ramp is not greater than a first distance threshold, the current perception fusion strategy is switched to a fusion strategy based on millimeter-wave radar, and a road passing through the ramp is output. According to the local path planning, the vehicle is controlled to enter the ramp and exit along the output road passing through the ramp.

2. The ramp driving assistance method according to claim 1, characterized in that: The method for outputting a road passing through the ramp using a millimeter-wave radar-based fusion strategy includes: Data alignment: Align the processed millimeter-wave radar road edges with the road edges of the high-precision map and the visual road edges to ensure that they are in the same coordinate system; Feature matching: performing feature matching on the road sidelines of the three data sources, and inspecting and correcting the detailed attributes of the processed millimeter-wave radar road sidelines; Road sideline polynomial fitting: performing cubic polynomial fitting on the corrected and processed millimeter-wave radar road sideline and outputting the fused road sideline; Virtualizing the lane line on the other side: based on the fused road edge line, virtualizing the lane line on the other side that is equal in width and parallel; Lane line detection output: compare and verify the virtual lane line on the other side with the road edge line of the high-precision map and the visual road edge line to ensure that the distance between the virtual lane line on the other side and the fused road edge line is smaller than the distance between the road edge line of the high-precision map and the visual road edge line and the fused road edge line, and output the virtual lane line on the other side.

3. The ramp driving assistance method according to claim 2, characterized in that: The processed millimeter-wave radar road edge line is obtained by identifying and processing the road edge line on the side of the direction in which the vehicle passes through the ramp based on the millimeter-wave radar, and the processing method includes: Data preprocessing: Obtain millimeter-wave radar point cloud data, use filtering algorithms to remove noise points, and separate ground points from non-ground points in the point cloud through cluster analysis; Feature extraction: identifying edge points in the point cloud that may belong to the road edge line, and extracting possible straight line segments from the edge points by linear regression method; Road sideline recognition: clustering the extracted straight line segments according to their direction, position and length features to distinguish the left and right sidelines, and confirming the road sideline on the direction in which the vehicle passes the ramp; Road model fitting: fitting the clustered straight line segments into a road edge model and verifying the accuracy of the road edge model; Post-processing: Smoothing the fitted road edge lines to eliminate small fluctuations and noise; using interpolation to fill in missing edge line data due to occlusion.

4. The ramp driving assistance method according to claim 2, characterized in that: The method for the vehicle to travel in the ramp is: In the local path planning through the ramp, a driving trajectory point set of the local path planning, as well as the lateral and longitudinal driving speeds and directions of the vehicle when the vehicle is located at the driving trajectory point set are planned at intervals of 1 m; wherein the driving trajectory point set is located on the center line of the fused road edge line and the virtual lane line on the other side.

5. The ramp driving assistance method according to claim 2, characterized in that: The width of the virtual equal-width, parallel lane lines on the other side is 3-3.2 meters.

6. The ramp driving assistance method according to claim 1, characterized in that: The method for calculating the deceleration distance is: Wherein, d is the deceleration distance; Δt is the time required for the vehicle to decelerate from the current speed to the speed limit of the ramp; v 车 is the current speed of the vehicle; v L is the speed limit of the ramp; a is the preset value of the vehicle's deceleration.

7. The ramp driving assistance method according to claim 1, characterized in that: The specific method for detecting whether there is a ramp within the first distance in front of the current position of the vehicle is: During driving, the vehicle detects whether there is a ramp within the first distance ahead of the current position of the vehicle in the global path at a preset frequency based on the high-precision map module in the intelligent driving domain controller; If the returned information is negative, the ramp detection loop continues; If the returned information is yes, then the ramp related information is obtained.

8. The ramp driving assistance method according to claim 7, characterized in that: The global path is the path planned by navigation-assisted driving between any two points within the range of the high-precision map.

9. The ramp driving assistance method according to claim 1, characterized in that: The relevant information of the ramp is the specific attribute information of the ramp on the high-precision map, including the geographic coordinates of the starting point and end point of the ramp, the ramp category, speed limit information, the geographic coordinates of the path guidance point, and the heading angle information; wherein the ramp category is divided into single-driving direction and dual-driving direction.

10. The ramp driving assistance method according to claim 1, characterized in that: The method for obtaining the direction of the vehicle passing through the ramp is: If the ramp type is single-direction, the direction of the vehicle passing through the ramp is to the right; If the ramp type is dual-direction, when a vehicle passes through the ramp from the right fork of the ramp, the direction of the vehicle passing through the ramp is to the right; When a vehicle passes through the ramp from the left fork of the ramp, the direction of the vehicle passing through the ramp is to the left; The local planning is that the direction in which the vehicle passes through the ramp is left or right.

Images