Intersection-crossing line selection processing method and device and automobile

Abstract

The invention provides an intersection-crossing line selection processing method and device and an automobile. The method comprises the following steps: acquiring vehicle information of a vehicle and lane information of a plurality of lanes to be selected every preset time; for each to-be-selected lane, generating a candidate center line corresponding to the to-be-selected lane according to the current position information of the vehicle and the position information of the to-be-selected lane; for each candidate center line, according to the current position information of the vehicle, the historical center line information and the obstacle information of the lane to be selected, determining the consistency cost value, the comfort cost value and the safety cost value of the candidate center line; the cost value of the candidate center line is determined according to the consistency cost value, the comfort cost value and the safety cost value; and determining the minimum value in the cost values of the plurality of candidate center lines, determining the candidate center line corresponding to the minimum value as the center line of the vehicle, and controlling the vehicle to run according to the center line. According to the method provided by the invention, the driving route with safety and comfort can be planned.

Description

Technical Field

[0001] This application relates to the field of driver assistance technology, and in particular to a method, device and vehicle for lane selection at intersections. Background Technology

[0002] Advanced driver assistance systems (ADAS), as technologies that provide support to drivers during driving, aim to provide safer and more comfortable driving routes in various driving scenarios. In scenarios involving crossing intersections, it is common to choose from multiple lanes on the opposite side of the intersection to plan a route.

[0003] In existing technologies, lane selection is typically based on high-precision maps, and driving routes are planned according to the selected lanes to guide vehicles through intersections. However, this method relies excessively on high-precision maps. If the high-precision maps are not updated in a timely manner, lanes that are impassable due to construction or road reconstruction may be selected, leading to incorrect driving route planning. Moreover, this method cannot detect obstacles in the lanes in real time, making it difficult to guarantee the safety and comfort of the planned driving routes. Summary of the Invention

[0004] This application provides a method, device, and vehicle for lane selection at intersections, which solves the problem that existing technologies, when selecting lanes at intersections, make it difficult to guarantee the safety and comfort of the planned driving route due to over-reliance on high-precision maps.

[0005] In a first aspect, this application provides a method for lane selection at intersections, comprising: acquiring vehicle information and lane information of multiple candidate lanes at preset time intervals; wherein the vehicle information includes the vehicle's current location information and historical centerline information, and the lane information includes the position information and obstacle information of the candidate lanes; for each candidate lane, generating a candidate centerline corresponding to the candidate lane based on the vehicle's current location information and the candidate lane's position information; for each candidate centerline, determining the consistency value, comfort value, and safety value of the candidate centerline based on the vehicle's current location information, historical centerline information, and obstacle information of the candidate lane, and determining the value of the candidate centerline based on the consistency value, comfort value, and safety value; determining the minimum value among the multiple candidate centerline values, and determining the candidate centerline corresponding to the minimum value as the vehicle's centerline, and controlling the vehicle to drive along the centerline.

[0006] In one specific implementation, determining the consistency cost of the candidate centerline based on the vehicle's historical centerline information includes: selecting multiple trajectory points on the candidate centerline and the historical centerline according to a preset step size; wherein the trajectory points on the candidate centerline correspond one-to-one with the trajectory points on the historical centerline; for each trajectory point on the candidate centerline, calculating the distance between the trajectory point and the corresponding trajectory point on the historical centerline; obtaining the weight value of each distance, and performing a weighted summation of the multiple distances based on the weight value of each distance to obtain the consistency cost of the candidate centerline.

[0007] In one specific implementation, determining the comfort value of the candidate centerline based on the vehicle's current location information includes: generating the vehicle's driving trajectory based on the vehicle's current location information and the candidate centerline; selecting multiple trajectory points on the driving trajectory according to a preset step size, and obtaining speed change information for each trajectory point; obtaining a weight value for each speed change information, and performing a weighted summation of the speed change information of the multiple trajectory points based on the weight value of each speed change information to obtain the comfort value of the candidate centerline.

[0008] In one specific implementation, determining the safety cost of the candidate centerline based on the vehicle's current location information and the obstacle information of the lane to be selected includes: generating the vehicle's driving trajectory based on the vehicle's current location information and the candidate centerline; obtaining the minimum distance between each obstacle on the lane to be selected corresponding to the candidate centerline and the driving trajectory, and determining the minimum value of a plurality of the minimum distances; and determining the safety cost of the candidate centerline based on the minimum value.

[0009] In one specific implementation, before determining the value of the candidate centerline based on the consistency cost value, comfort cost value, and safety cost value, the method further includes: obtaining a consistency cost threshold for the consistency cost value, a comfort cost threshold for the comfort cost value, and a safety cost threshold for the safety cost value, respectively; then, determining the value of the candidate centerline based on the consistency cost value, comfort cost value, and safety cost value includes: when it is determined that the consistency cost value is less than the consistency cost threshold, the comfort cost value is less than the comfort cost threshold, and the safety cost value is less than the safety cost threshold, determining the value of the candidate centerline based on the consistency cost value, comfort cost value, and safety cost value.

[0010] In one specific embodiment, the method further includes: acquiring navigation information of the vehicle, lane line information and traffic flow speed information of multiple candidate lanes, and at least one of the historical driving trajectory and current driving information of vehicles adjacent to the vehicle; determining at least one of the road trend value, traffic flow value, navigation value, and driving efficiency value of the candidate centerline based on the navigation information of the vehicle, lane line information and traffic flow speed information of multiple candidate lanes, and at least one of the historical driving trajectory and current driving information of vehicles adjacent to the vehicle; the step of determining the value of the candidate centerline based on the consistency value, comfort value, and safety value includes: determining the value of the candidate centerline based on the consistency value, comfort value, safety value, and at least one of the road trend value, traffic flow value, navigation value, and driving efficiency value.

[0011] In one specific implementation, determining the road trend value of the candidate centerline based on lane line information of multiple candidate lanes includes: performing rasterization processing on the candidate lanes corresponding to the candidate centerline based on lane line information of multiple candidate lanes, and obtaining the raster direction of each candidate lane raster; selecting multiple trajectory points on the candidate centerline according to a preset step size; at each trajectory point, obtaining the angle between the candidate centerline and the raster direction of the candidate lane raster where the trajectory point is located; obtaining the weight value of each angle, and performing weighted summation processing on multiple angles based on the weight value of each angle to obtain the road trend value of the candidate centerline; and / or, based on the relationship with the vehicle... The traffic flow value of the candidate centerline is determined by analyzing the historical driving trajectories and current driving information of neighboring vehicles. This includes: generating predicted lanes for vehicles adjacent to the vehicle based on their historical driving trajectories and current driving information; performing rasterization on the predicted lanes and obtaining the raster orientation of each predicted lane raster; selecting multiple trajectory points on the candidate centerline according to a preset step size; obtaining the angle between the candidate centerline and the raster orientation of the predicted lane raster corresponding to each trajectory point; obtaining the weight value of each angle; and performing a weighted summation of multiple angles based on the weight value of each angle to obtain the traffic flow value of the candidate centerline.

[0012] In one specific implementation, determining the navigation cost of the candidate centerline based on the vehicle's navigation information includes: determining the number of navigation lane changes and the remaining drivable distance required for the vehicle in the candidate lane corresponding to the candidate centerline based on the vehicle's navigation information; obtaining the weight values ​​corresponding to the number of navigation lane changes and the remaining drivable distance, respectively; performing a weighted summation of the number of navigation lane changes and the remaining drivable distance based on the weight values ​​corresponding to the number of navigation lane changes and the remaining drivable distance to obtain the navigation cost of the candidate centerline; and / or, determining the driving efficiency cost of the candidate centerline based on the traffic flow speed information of multiple candidate lanes includes: determining the traffic flow speed of the vehicle in the candidate lane corresponding to the candidate centerline based on the traffic flow speed information of multiple candidate lanes; and determining the driving efficiency cost of the candidate centerline based on the traffic flow speed.

[0013] In a second aspect, this application provides an electronic device, including: a processor, a memory, and a communication interface; the memory is used to store executable instructions of the processor; wherein the processor is configured to execute the intersection lane selection processing method described in the first aspect by executing the executable instructions.

[0014] Thirdly, this application provides an automobile, including: electronic equipment as described in the second aspect.

[0015] This application provides a method, device, and vehicle for lane selection at intersections. The method includes acquiring vehicle information and lane information of multiple candidate lanes at preset time intervals; for each candidate lane, generating a candidate centerline corresponding to the candidate lane based on the vehicle's current location information and the candidate lane's location information; for each candidate centerline, determining the consistency value, comfort value, and safety value of the candidate centerline based on the vehicle's current location information, historical centerline information, and obstacle information of the candidate lane, and determining the value of the candidate centerline based on the consistency value, comfort value, and safety value; determining the minimum value among the multiple candidate centerline values, and determining the candidate centerline corresponding to the minimum value as the vehicle's centerline, controlling the vehicle to travel along the centerline. The method of this application can evaluate the consistency value, comfort value, and safety value of the centerline of each candidate lane based on vehicle information and lane information of multiple candidate lanes, so as to determine the candidate lane corresponding to the optimal centerline, thereby planning a driving route with safety and comfort. This solves the problem that the existing technology makes it difficult to guarantee the safety and comfort of the planned driving route when selecting lanes at intersections due to over-reliance on high-precision maps. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0017] Figure 1 A flowchart illustrating an embodiment of the alignment selection method for intersections provided in this application;

[0018] Figure 2 A flowchart illustrating a second embodiment of the alignment selection method for intersections provided in this application;

[0019] Figure 3 A schematic diagram illustrating the consistency cost value for determining candidate centerlines provided in this application;

[0020] Figure 4 A flowchart illustrating a third embodiment of the alignment selection method for intersections provided in this application;

[0021] Figure 5 A schematic diagram illustrating the comfort cost values ​​for determining candidate centerlines provided in this application;

[0022] Figure 6 A schematic flowchart of Embodiment 4 of the intersection selection processing method provided in this application;

[0023] Figure 7 A schematic diagram illustrating the security cost value for determining candidate centerlines provided in this application;

[0024] Figure 8 A flowchart illustrating Embodiment 5 of the alignment selection method for intersections provided in this application;

[0025] Figure 9 A flowchart illustrating Embodiment Six of the intersection selection method provided in this application;

[0026] Figure 10 A flowchart illustrating Embodiment Seven of the intersection selection method provided in this application;

[0027] Figure 11 A schematic diagram illustrating the road trend cost value for determining candidate centerlines provided in this application;

[0028] Figure 12 A flowchart illustrating an eighth embodiment of a method for lane selection at intersections provided in this application;

[0029] Figure 13 A schematic diagram illustrating the traffic cost value for determining the candidate centerline provided in this application;

[0030] Figure 14 A flowchart illustrating Embodiment Nine of the intersection selection processing method provided in this application;

[0031] Figure 15 A flowchart illustrating Embodiment 10 of the method for lane selection at intersections provided in this application;

[0032] Figure 16 A schematic diagram of the structure of an embodiment of a lane selection processing device for an intersection provided in this application;

[0033] Figure 17 This is a schematic diagram of the structure of an embodiment of an electronic device provided in this application.

[0034] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0035] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0036] Advanced driver assistance systems (ADAS), as technologies that provide support to drivers during driving, aim to provide safer and more comfortable driving routes in various driving scenarios. In scenarios involving crossing intersections, it is common to choose from multiple lanes on the opposite side of the intersection to plan a route.

[0037] In existing technologies, lane selection is typically based on high-precision maps, and driving routes are planned according to the selected lanes to guide vehicles through intersections. However, this method relies excessively on high-precision maps. If the high-precision maps are not updated in a timely manner, lanes that are impassable due to construction or road reconstruction may be selected, leading to incorrect driving route planning. Moreover, this method cannot detect obstacles in the lanes in real time, making it difficult to guarantee the safety and comfort of the planned driving routes.

[0038] Based on the above-mentioned technical problems, the technical concept of this application is as follows: How to provide a lane selection processing method for crossing intersections, which can determine the optimal lane from multiple candidate lanes when a vehicle crosses an intersection, so as to plan a driving route with safety and comfort.

[0039] The method provided in this application is intended to solve the above-mentioned technical problems of the prior art.

[0040] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0041] Figure 1 This is a flowchart illustrating an embodiment of a lane selection method for intersections provided in this application. See also... Figure 1 The lane selection process at this intersection can be performed by a lane selection device or electronic equipment at the intersection, and specifically includes the following steps:

[0042] Step S101: At preset intervals, obtain vehicle information and lane information of multiple candidate lanes.

[0043] The vehicle information includes the vehicle's current location and historical centerline information, while the lane information includes the location information of the candidate lane and obstacle information.

[0044] In this embodiment, vehicle information and lane information of multiple candidate lanes can be acquired at preset time intervals. For example, vehicle information and lane information of multiple candidate lanes can be acquired periodically starting when the vehicle approaches an intersection. For instance, vehicle information and lane information can be acquired every 200ms starting when the vehicle is 200 meters before the intersection.

[0045] Specifically, vehicle information may include the vehicle's current location information and historical centerline information. The historical centerline is the optimal centerline among multiple candidate centerlines determined at the previous moment, and is updated when a new optimal centerline is determined at the next moment. For example, the vehicle's current location information may include the vehicle's current latitude and longitude information, as well as the lane information the vehicle is currently in. The vehicle's historical centerline information may include the trajectory information of the vehicle's centerline at the previous moment. The vehicle's current location information can be obtained through the vehicle's positioning device and / or navigation device, while the vehicle's historical centerline information can be obtained from local storage.

[0046] Lane information may include the location information and obstacle information of the candidate lane. For example, the location information of the candidate lane may include its latitude and longitude, exit information, etc. The obstacle information of the candidate lane may include static and dynamic obstacle information included in the candidate lane. The location information and obstacle information of the candidate lane can be obtained based on images acquired using visual perception technology. For example, surrounding images acquired by visual acquisition devices installed on the vehicle or roadside visual acquisition devices communicating with the vehicle can be used for image recognition to obtain the location information and obstacle information of the candidate lane.

[0047] Step S102: For each candidate lane, generate a candidate centerline corresponding to the candidate lane based on the current location information of the vehicle and the location information of the candidate lane.

[0048] In this embodiment, for each candidate lane, a candidate centerline can be generated based on the vehicle's current location information and the lane's location information. For example, the candidate centerline can be generated using an Iterative Linear Quadratic Regulator (iLQR) based on the vehicle's current lane information and the candidate lane's exit information. Alternatively, the vehicle's current lane and the candidate lane's exit can be connected, and a smooth curve fitted using a polynomial curve can be obtained as the candidate centerline. Thus, multiple candidate centerlines can be generated.

[0049] Step S103: For each candidate centerline, based on the vehicle's current location information, historical centerline information, and obstacle information of the candidate lane, determine the consistency value, comfort value, and safety value of the candidate centerline, and determine the value of the candidate centerline based on the consistency value, comfort value, and safety value.

[0050] In this embodiment, the cost value of each candidate centerline can be determined. Specifically, the cost value of a candidate centerline can be determined based on its consistency cost value, comfort cost value, and safety cost value. The consistency cost value represents the degree of consistency between the candidate centerline and historical centerlines; a higher degree of consistency results in a lower consistency cost value. The comfort cost value represents the level of comfort of the vehicle traveling along the candidate centerline; a higher level of comfort results in a lower comfort cost value. The safety cost value represents the level of safety of the vehicle traveling along the candidate centerline; a higher level of safety results in a lower safety cost value.

[0051] In one possible implementation, the consistency value of a candidate centerline can be determined based on the vehicle's historical centerline information. For example, multiple trajectory points can be selected on both the candidate and historical centerlines, and the distances between these points can be calculated. These distances are then weighted and summed to obtain the consistency value of the candidate centerline. The distances between the trajectory points can be lateral, longitudinal, or Euclidean distances, etc. Thus, the consistency of each candidate centerline can be evaluated based on the vehicle's historical centerline information. The closer the candidate centerline is to the historical centerline, the higher the degree of consistency and the lower the consistency value.

[0052] In one possible implementation, the comfort value of a candidate centerline can be determined based on the vehicle's current location information. For example, a driving trajectory for the vehicle can be generated based on its current location information and the candidate centerlines. Multiple trajectory points can be selected along this trajectory, and the speed change information at each point can be obtained. The speed change information from multiple trajectory points can then be weighted and summed to obtain the comfort value of the candidate centerline. Thus, the comfort of each candidate centerline can be evaluated based on the vehicle's current location information. The smaller the speed change when the vehicle travels along the candidate centerline, the higher the comfort level and the lower the comfort value.

[0053] In one possible implementation, the safety cost of a candidate centerline can be determined based on the vehicle's current location information and obstacle information of the lane to be selected. For example, based on the vehicle's current location information and the candidate centerlines, a driving trajectory of the vehicle can be generated, the minimum distance between each obstacle in the lane to be selected and the driving trajectory can be obtained, and the minimum value of multiple minimum distances can be determined; based on this minimum value, the safety cost of the candidate centerline can be determined. Thus, the safety of each candidate centerline can be evaluated based on the vehicle's current location information and obstacle information of the lane to be selected. When the vehicle travels along a candidate centerline, the greater the distance from surrounding obstacles, the higher the safety level and the lower the safety cost.

[0054] In one possible implementation, the weight values ​​a1, a2, and a3 corresponding to the consistency value, comfort value, and safety value can be obtained separately. Based on these weight values, a1, a2, and a3, a weighted sum is applied to the consistency value A1, comfort value A2, and safety value A3 to obtain the value A of the candidate centerline: A = a1*A1 + a2*A2 + a3*A3. Therefore, the consistency, comfort, and safety values ​​of each candidate centerline can be comprehensively evaluated.

[0055] Frequent fluctuations in the center line cause vehicles to frequently change direction when crossing intersections, affecting driving experience and safety. Therefore, the consistency between the candidate center line and the historical center line is a more important consideration when selecting a candidate center line, and vehicle safety is more important than comfort. Thus, in one possible implementation, the weight value corresponding to consistency can be the highest, followed by the weight value corresponding to safety, and then the weight value corresponding to comfort. For example, the weight value a1 corresponding to consistency > the weight value a3 corresponding to safety > the weight value a2 corresponding to comfort.

[0056] Step S104: Determine the minimum value among multiple candidate centerlines, and determine the candidate centerline corresponding to the minimum value as the centerline of the vehicle, and control the vehicle to travel along the centerline.

[0057] In this embodiment, after evaluating the consistency, comfort, and safety values ​​of each candidate centerline, the minimum value among the multiple candidate centerlines can be determined. The candidate centerline corresponding to this minimum value is the one with the highest consistency, comfort, and safety among the multiple candidate centerlines. The candidate centerline corresponding to this minimum value can be determined as the vehicle's centerline, and the vehicle can be controlled to travel along this centerline.

[0058] In this embodiment, vehicle information and lane information of multiple candidate lanes are acquired at preset time intervals. For each candidate lane, a candidate centerline is generated based on the vehicle's current location information and the lane's location information. For each candidate centerline, the consistency value, comfort value, and safety value are determined based on the vehicle's current location information, historical centerline information, and obstacle information of the lane. The value of the candidate centerline is then determined based on these values. The minimum value among the multiple candidate centerlines is determined, and the candidate centerline corresponding to this minimum value is identified as the vehicle's centerline. The vehicle is then controlled to travel along this centerline. The method of this application can evaluate the consistency value, comfort value, and safety value of the centerline of each candidate lane based on vehicle information and lane information of multiple candidate lanes, so as to determine the candidate lane corresponding to the optimal centerline, thereby planning a driving route with safety and comfort. This solves the problem that the existing technology makes it difficult to guarantee the safety and comfort of the planned driving route when selecting lanes at intersections due to over-reliance on high-precision maps.

[0059] Figure 2 This is a flowchart illustrating a second embodiment of the intersection lane selection method provided in this application. See also... Figure 2 The above step S103 specifically includes the following steps:

[0060] Step S201: Select multiple trajectory points on the candidate center line and the historical center line according to the preset step size.

[0061] The trajectory points on the candidate center line correspond one-to-one with the trajectory points on the historical center line.

[0062] In this embodiment, multiple trajectory points can be selected on both the candidate centerline and the historical centerline according to a preset step size. For example, the preset step size can be 2 meters. The historical centerline can be the centerline from the previous moment. Figure 3 A schematic diagram illustrating the consistency cost value for determining candidate centerlines provided in this application. Figure 3 As shown, multiple trajectory points p1, p2, p3, and p4 can be selected on the candidate centerline starting from the starting point or from a preset distance from the starting point, according to a preset step size; similarly, multiple trajectory points p'1, ​​p'2, p'3, and p'4 can be selected on the historical centerline starting from the starting point or from a preset distance from the starting point. Here, p1 corresponds to p'1, ​​p2 corresponds to p'2, p3 corresponds to p'3, and p4 corresponds to p'4. For example, multiple trajectory points on the historical centerline can be obtained by projecting multiple trajectory points on the candidate centerline; for instance, p'1 is obtained by projecting p1 onto the historical centerline, p'2 is obtained by projecting p2 onto the historical centerline, p'3 is obtained by projecting p3 onto the historical centerline, and p'4 is obtained by projecting p4 onto the historical centerline. In one possible implementation, multiple corresponding trajectory points can be selected on the candidate centerline and the historical centerline according to a preset number of trajectory points and the rule that the closer to the current position of the vehicle, the shorter the step size, and the farther away from the current position of the vehicle, the longer the step size.

[0063] In one possible implementation, multiple trajectory points can be located before entering the intersection, during the intersection, and after passing the intersection, respectively.

[0064] Step S202: For each trajectory point on the candidate center line, calculate the distance between the trajectory point and the corresponding trajectory point on the historical center line.

[0065] In this embodiment, for each trajectory point p1, p2, p3, and p4 on the candidate center line, the distance between that trajectory point and the corresponding trajectory point on the historical center line can be calculated. For example... Figure 3 As shown, the distances d1 between p1 and p'1, ​​d2 between p2 and p'2, d3 between p3 and p'3, and d4 between p4 and p'4 can be calculated. The distances between trajectory points can be the horizontal distance on the x-axis, the vertical distance on the y-axis, or the Euclidean distance, etc.

[0066] Step S203: Obtain the weight value of each distance, and perform a weighted summation of multiple distances based on the weight value of each distance to obtain the consistency cost of the candidate centerline.

[0067] In this embodiment, multiple distances can be weighted and summed to obtain the consistency cost of the candidate centerline. Specifically, the weight value of each distance can be obtained, and multiple distances can be weighted and summed based on the weight value of each distance. As mentioned in the previous example, multiple distances d1, d2, d3, d4 are weighted and summed based on the weight values ​​k1, k2, k3, k4 of each distance to obtain the consistency cost A1 of the candidate centerline = k1*d1 + k2*d2 + k3*d3 + k4*d4. For example, the closer the trajectory point is to the current position of the vehicle, the larger the weight value of the distance.

[0068] In this embodiment, the consistency of the candidate centerline is evaluated by calculating the distance between trajectory points on the candidate centerline and trajectory points on the historical centerline, and then performing a weighted summation of multiple distances. This provides a prerequisite for subsequently selecting candidate centerlines with lower consistency costs. Therefore, the centerline can be made as consistent as possible with the historical centerline, avoiding frequent centerline jumps that cause vehicles to frequently change direction when crossing intersections, thus affecting driving experience and safety.

[0069] Figure 4 This is a flowchart illustrating a third embodiment of the intersection lane selection method provided in this application. See also... Figure 4 The above step S103 specifically includes the following steps:

[0070] Step S401: Generate the vehicle's driving trajectory based on the vehicle's current location information and candidate center lines.

[0071] In this embodiment, the vehicle's driving path and speed can be planned using an iterative linear quadratic regulator iLQR based on the vehicle's current location information and candidate centerline, thereby generating the vehicle's driving trajectory, which is the trajectory of the vehicle traveling along the candidate centerline based on its current location. Figure 5 A schematic diagram illustrating the comfort cost values ​​for determining candidate centerlines provided in this application.

[0072] Step S402: Select multiple trajectory points on the driving trajectory according to the preset step size, and obtain the speed change information of each trajectory point.

[0073] In this embodiment, multiple trajectory points can be selected on the driving trajectory according to a preset step size. For example, the preset step size can be 2 meters. Figure 5 As shown, multiple trajectory points q1, q2, q3, q4, q5 and q6 are selected on the driving trajectory according to the preset step size.

[0074] In one possible implementation, multiple trajectory points can be selected on the driving trajectory according to a preset number of trajectory points and the rule that the closer to the vehicle's current position, the shorter the step length, and the farther away from the vehicle's current position, the longer the step length.

[0075] In one possible implementation, multiple trajectory points can be located before entering the intersection, during the intersection, and after passing the intersection, respectively.

[0076] The velocity change information for each trajectory point can be obtained separately. This velocity change information can be the vehicle's lateral acceleration, longitudinal acceleration, lateral jerk, or longitudinal jerk at that trajectory point.

[0077] Step S403: Obtain the weight value of each velocity change information, and perform weighted summation of the velocity change information of multiple trajectory points according to the weight value of each velocity change information to obtain the comfort cost of the candidate centerline.

[0078] In this embodiment, the speed change information of multiple trajectory points can be weighted and summed to obtain the comfort value of the candidate centerline. Specifically, the weight value of each speed change information can be obtained, and the speed change information of multiple trajectory points can be weighted and summed based on the weight value of each speed change information. As mentioned in the previous example, the speed change information j1, j2, j3, m4, m5, and j6 of multiple trajectory points are weighted and summed based on the weight values ​​m1, m2, m3, m4, m5, and m6 of each speed change information to obtain the comfort value A2 of the candidate centerline = m1*j1 + m2*j2 + m3*j3 + m4*j4 + m5*j5 + m6*j6. For example, the closer the trajectory point is to the current position of the vehicle, the larger the weight value of the speed change information.

[0079] In this embodiment, the comfort of candidate centerlines is evaluated by acquiring the speed change information of each trajectory point on the driving trajectory and performing a weighted summation of the speed change information of multiple trajectory points. This provides a prerequisite for subsequently selecting candidate centerlines with lower comfort costs. Therefore, candidate centerlines corresponding to driving trajectories with smaller speed changes can be selected, improving driving comfort.

[0080] Figure 6 This is a flowchart illustrating Embodiment 4 of the intersection lane selection method provided in this application. See also... Figure 6 The above step S103 specifically includes the following steps:

[0081] Step S601: Generate the vehicle's driving trajectory based on the vehicle's current location information and candidate center lines.

[0082] In this embodiment, for each candidate centerline, a vehicle trajectory can be generated based on the vehicle's current location information and the candidate centerline. The trajectory is the path along the candidate centerline based on the vehicle's current location. Figure 7 A schematic diagram illustrating the security cost value for determining candidate centerlines provided in this application. (See diagram below.) Figure 7 As shown, when a vehicle crosses an intersection, there are three candidate lanes, each corresponding to a candidate centerline. Based on the vehicle's current location information and the three candidate centerlines, three driving trajectories for the vehicle are generated: driving trajectory 1, driving trajectory 2, and driving trajectory 3.

[0083] Step S602: Obtain the minimum distance between each obstacle on the candidate lane corresponding to the candidate centerline and the driving trajectory, and determine the minimum value of multiple minimum distances.

[0084] In this embodiment, for each driving trajectory, the minimum distance between each obstacle on the candidate lane corresponding to the candidate centerline that generated the driving trajectory and the driving trajectory can be obtained. For example... Figure 7 As shown, multiple trajectory points c1, c2, c3, and c4 can be selected on the driving trajectory 3 according to a preset step size. The distance between each trajectory point and an obstacle on the candidate lane is calculated, and the minimum value among the multiple distances is determined. This minimum value is the minimum distance between the obstacle and the driving trajectory. A minimum distance is generated for each obstacle, and multiple obstacles will generate multiple minimum distances. For example, the distance between the trajectory point and the obstacle can be the lateral distance on the horizontal axis, the longitudinal distance on the vertical axis, or the Euclidean distance, etc.

[0085] In one possible implementation, for each driving trajectory, the distance between the driving trajectory and the position coordinates of each obstacle on the candidate lane can be calculated based on the position coordinates of each obstacle, and the minimum value of multiple distances can be determined.

[0086] Specifically, when the obstacle on the candidate lane is a dynamic obstacle, for each driving trajectory, the driving trajectory and the running trajectory of the dynamic obstacle can be aligned according to the timestamp, the distance between the driving trajectory and the running trajectory of the dynamic obstacle can be calculated, and the minimum value among multiple distances can be determined. This minimum value is the minimum distance between the dynamic obstacle and the driving trajectory.

[0087] For each driving trajectory, when there are multiple obstacles in the candidate lane, the minimum distance between each obstacle in the candidate lane and the driving trajectory can be obtained separately, and the minimum value of multiple minimum distances can be determined.

[0088] Step S603: Determine the safety cost of the candidate centerline based on the minimum value.

[0089] In this embodiment, the safety cost of a candidate centerline can be determined based on the minimum value of multiple minimum distances. For example, a correspondence between the minimum value and the safety cost A3 can be set, or a proportional coefficient can be set so that the minimum value and the safety cost are inversely proportional, with the smaller the minimum value, the greater the safety cost.

[0090] In this embodiment, the safety of candidate centerlines is evaluated by obtaining the minimum distance between each obstacle in the candidate lane and the driving trajectory, and selecting the minimum value from multiple minimum distances. This provides a prerequisite for subsequently screening candidate centerlines with lower safety costs. Therefore, candidate centerlines corresponding to driving trajectories with larger distances from obstacles in the candidate lane can be selected, improving driving safety.

[0091] Figure 8 This is a flowchart illustrating Embodiment 5 of the intersection lane selection method provided in this application. See also... Figure 8 The lane selection process at this intersection can be performed by a lane selection device or electronic equipment at the intersection, and specifically includes the following steps:

[0092] Step S801: At preset intervals, obtain vehicle information and lane information of multiple candidate lanes.

[0093] The vehicle information includes the vehicle's current location and historical centerline information, while the lane information includes the location information of the candidate lane and obstacle information.

[0094] In this embodiment, vehicle information and lane information of multiple candidate lanes can be acquired at preset time intervals. For example, vehicle information and lane information of multiple candidate lanes can be acquired periodically starting when the vehicle approaches an intersection. Surrounding images acquired by a vision acquisition device installed on the vehicle or a roadside vision acquisition device communicating with the vehicle can be used for image recognition to obtain the location information of the candidate lanes, thereby determining the candidate lanes.

[0095] Step S802: For each candidate lane, generate a candidate centerline corresponding to the candidate lane based on the current location information of the vehicle and the location information of the candidate lane.

[0096] In this embodiment, for each candidate lane, a candidate centerline can be generated based on the vehicle's current location information and the lane's location information. For example, the candidate centerline can be generated based on the vehicle's current lane information and the lane's exit information. Thus, multiple candidate centerlines can be generated.

[0097] Step S803: For each candidate centerline, based on the vehicle's current location information, historical centerline information, and obstacle information of the candidate lane, determine the consistency value, comfort value, and safety value of the candidate centerline respectively.

[0098] In this embodiment, the consistency value of the candidate centerline can be determined based on the vehicle's historical centerline information; the comfort value of the candidate centerline can be determined based on the vehicle's current location information; and the safety value of the candidate centerline can be determined based on the vehicle's current location information and the obstacle information of the lane to be selected.

[0099] Step S804: Obtain the consistency cost threshold of the consistency cost value, the comfort cost threshold of the comfort cost value, and the safety cost threshold of the safety cost value, respectively.

[0100] In this embodiment, after determining the consistency cost, comfort cost, and safety cost of the candidate centerline, the consistency cost threshold, comfort cost threshold, and safety cost threshold can be obtained respectively. These thresholds can be pre-set and used to filter out candidate centerlines with unreasonable cost values.

[0101] Taking safety cost as an example, if the safety cost of a candidate centerline is greater than the safety cost threshold, then when the vehicle travels along the candidate centerline, the distance to surrounding obstacles is too small, posing a collision risk. Therefore, the candidate centerline is an unreasonable centerline.

[0102] In one possible implementation, the consistency cost threshold, comfort cost threshold, and safety cost threshold of the consistency cost value are different, and the specific thresholds can be determined according to actual requirements or experience.

[0103] Step S805: When it is determined that the consistency cost value is less than the consistency cost threshold, the comfort cost value is less than the comfort cost threshold, and the safety cost value is less than the safety cost threshold, the cost value of the candidate centerline is determined based on the consistency cost value, the comfort cost value, and the safety cost value.

[0104] In this embodiment, when the consistency cost value of a candidate centerline is less than the consistency cost threshold, the comfort cost value is less than the comfort cost threshold, and the safety cost value is less than the safety cost threshold, the cost value of the candidate centerline can be determined to be reasonable. The cost value of the candidate centerline can be determined based on the consistency cost value, the comfort cost value, and the safety cost value.

[0105] In one possible implementation, the weight values ​​corresponding to the consistency cost, comfort cost, and safety cost can be obtained separately. Based on these weight values, a weighted sum is applied to the consistency cost, comfort cost, and safety cost to obtain the cost value of the candidate centerline. Therefore, the consistency, comfort, and safety of each candidate centerline can be evaluated by comprehensively considering their respective costs.

[0106] Step S806: Determine the minimum value among multiple candidate centerlines, and determine the candidate centerline corresponding to the minimum value as the centerline of the vehicle, and control the vehicle to travel along the centerline.

[0107] In this embodiment, after evaluating the consistency, comfort, and safety values ​​of each candidate centerline, the minimum value among the multiple candidate centerlines can be determined. The candidate centerline corresponding to this minimum value is the one with the highest consistency, comfort, and safety among the multiple candidate centerlines. The candidate centerline corresponding to this minimum value can be determined as the vehicle's centerline, and the vehicle can be controlled to travel along this centerline.

[0108] In this embodiment, by filtering out candidate centerlines with unreasonable consistency cost, comfort cost, and safety cost values ​​based on the consistency cost threshold, comfort cost threshold, and safety cost threshold, the candidate lanes corresponding to the optimal centerline can be determined more accurately, thereby planning a driving route that is both safe and comfortable.

[0109] Figure 9 A flowchart illustrating Embodiment Six of the intersection lane selection method provided in this application. See also... Figure 9 The alignment selection method for this intersection also includes the following steps:

[0110] Step S901: Obtain the vehicle's navigation information, lane line information and traffic speed information of multiple candidate lanes, and at least one of the historical driving trajectory and current driving information of vehicles adjacent to the vehicle.

[0111] In this embodiment, at least one of the following can be acquired: vehicle navigation information, lane line information and traffic speed information of multiple candidate lanes, and historical driving trajectories and current driving information of vehicles adjacent to the vehicle. For example, this information can be acquired periodically starting when the vehicle approaches an intersection. The vehicle navigation information can be obtained from a map database or online map data. The lane line information, traffic speed information, and current driving information of vehicles adjacent to the vehicle can be obtained by identifying surrounding images captured by a vision acquisition device installed on the vehicle or a roadside vision acquisition device communicating with the vehicle. The historical driving trajectories of vehicles adjacent to the vehicle can be obtained from locally stored images captured by the vision acquisition device.

[0112] Specifically, the vehicle's navigation information may include lane change information from the current lane to each candidate lane and the distance information of the vehicle from the intersection; the lane line information of the candidate lane may include the position and length information of the lane line; the traffic flow speed information of the candidate lane may include the average speed of vehicles traveling in the candidate lane; the vehicles adjacent to the vehicle may include vehicles traveling in adjacent lanes; the historical driving trajectory may include the driving trajectory information of the vehicles adjacent to the vehicle at the previous moment; and the current driving information may include the current driving direction information of the vehicles adjacent to the vehicle.

[0113] Step S902: Based on the vehicle's navigation information, lane line information and traffic flow speed information of multiple candidate lanes, and at least one of the historical driving trajectory and current driving information of vehicles adjacent to the vehicle, determine at least one of the road trend value, traffic flow value, navigation value and driving efficiency value of the candidate centerline.

[0114] The candidate centerline's road trend value represents the degree to which a vehicle traveling along the candidate centerline conforms to the current road trend; the higher the conformity, the lower the road trend value. The candidate centerline's traffic flow value represents the probability that a vehicle traveling along the candidate centerline will share a lane with other vehicles; the lower the probability, the lower the traffic flow value. The candidate centerline's navigation value represents the degree to which a vehicle traveling along the candidate centerline matches the current navigation information; the higher the matching degree, the lower the navigation value. The candidate centerline's driving efficiency value represents the driving efficiency of a vehicle traveling along the candidate centerline; the higher the driving efficiency, the lower the driving efficiency value.

[0115] In this embodiment, the road trend value of a candidate centerline can be determined based on lane line information of multiple candidate lanes. For example, the candidate lanes corresponding to the candidate centerline can be rasterized. Multiple trajectory points are selected on the candidate centerline, and at each trajectory point, the angle between the candidate centerline and the grid direction of the candidate lane grid where the trajectory point is located is obtained. The multiple angles are then weighted and summed to obtain the road trend value of the candidate centerline. When a vehicle travels along the candidate centerline, the smaller the angle between each trajectory point and the grid direction, the higher the degree of conformity with the current road trend, and the lower the road trend value.

[0116] The traffic flow cost of a candidate centerline can be determined based on the historical driving trajectories and current driving information of vehicles adjacent to the current vehicle. For example, a predicted lane for vehicles adjacent to the current vehicle can be generated, and this predicted lane can be rasterized. Multiple trajectory points are selected on the candidate centerline, and at each trajectory point, the angle between the candidate centerline and the grid direction of the predicted lane grid containing that trajectory point is obtained. These angles are then weighted and summed to obtain the traffic flow cost of the candidate centerline. When a vehicle travels along the candidate centerline, the smaller the angle between each trajectory point and the grid direction, the lower the probability of it sharing the same lane with other vehicles, and the lower the traffic flow cost.

[0117] The navigation cost of a candidate centerline can be determined based on the vehicle's navigation information. For example, the required number of lane changes and the remaining drivable distance required for the vehicle to navigate to the candidate lane corresponding to the centerline can be determined based on the vehicle's navigation information. A weighted sum of these factors yields the navigation cost of the candidate centerline. When a vehicle travels along the candidate centerline, fewer lane changes, a longer remaining drivable distance, and a higher degree of conformity with the current navigation information result in a lower navigation cost.

[0118] The driving efficiency cost of a candidate centerline can be determined based on the traffic speed information of multiple candidate lanes. For example, the traffic speed of a vehicle in the candidate lane corresponding to the candidate centerline can be determined based on the traffic speed information of multiple candidate lanes, and the driving efficiency cost of the candidate centerline can be determined based on this traffic speed. When a vehicle travels along the candidate centerline, the higher the traffic speed, the higher the driving efficiency, and the lower the driving efficiency cost.

[0119] The above step S103 specifically includes the following steps:

[0120] Step S903: Determine the value of the candidate centerline based on at least one of consistency value, comfort value, safety value, road trend value, traffic flow value, navigation value, and driving efficiency value.

[0121] In one possible implementation, at least one of the following weighted values ​​can be obtained: the weighted value corresponding to consistency, the weighted value corresponding to comfort, the weighted value corresponding to safety, and the weighted values ​​corresponding to road trend, traffic flow, navigation, and driving efficiency. Based on at least one of these weighted values, a weighted sum is applied to the consistency, comfort, and safety values, as well as the road trend, traffic flow, navigation, and driving efficiency values, to obtain the value of the candidate centerline. Therefore, each candidate centerline can be evaluated by comprehensively considering its consistency, comfort, and safety values, as well as at least one of these values.

[0122] In this embodiment, each candidate centerline can be evaluated by combining the consistency value, comfort value, and safety value of the candidate centerline, as well as at least one of the road trend value, traffic flow value, navigation value, and driving efficiency value. This allows for the determination of the candidate lane corresponding to the optimal centerline by incorporating more comfort and safety evaluation perspectives, thus planning a safer and more comfortable driving route.

[0123] Figure 10 This is a flowchart illustrating Embodiment Seven of the intersection lane selection method provided in this application. See also... Figure 10 The above step S902 specifically includes the following steps:

[0124] Step S1001: Based on the lane line information of multiple candidate lanes, perform rasterization processing on the candidate lanes corresponding to the candidate center lines, and obtain the raster direction of each candidate lane raster.

[0125] In this embodiment, the candidate lane corresponding to the candidate center line can be rasterized based on the lane line information of the candidate lane. The lane line information may include the position information and length information of the lane line. Figure 11 A schematic diagram illustrating the road trend cost values ​​for determining candidate centerlines provided in this application. (See diagram below.) Figure 11 As shown, the two candidate centerlines for vehicles crossing the intersection are indicated by solid lines, and each candidate centerline corresponds to a lane to be selected. The lane lines of the lanes to be selected are indicated by dashed lines. The lanes to be selected are rasterized, and the raster orientation of each lane raster is obtained, as shown by the arrows. For example, the raster orientation can be determined based on the lane line orientation.

[0126] Step S1002: Select multiple trajectory points on the candidate center line according to the preset step size.

[0127] In this embodiment, multiple trajectory points can be selected on the candidate centerline according to a preset step size. For example, the preset step size can be 2 meters. Figure 11 As shown, multiple trajectory points s1, s2, s3, and s4 can be selected on the candidate centerline according to a preset step size. For example, the multiple trajectory points can be located before entering the intersection, during the intersection, and after passing the intersection, respectively. Figure 11 As shown, trajectory point s1 is located before entering the intersection, s2 and s3 are located in the middle of the intersection, and s4 is located after passing the intersection.

[0128] Step S1003: At each trajectory point, obtain the angle between the candidate centerline and the grid direction of the candidate lane grid where the trajectory point is located.

[0129] In this embodiment, at each trajectory point, the angle between the candidate centerline and the grid direction of the candidate lane grid where that trajectory point is located can be obtained. This angle can be the angle between the tangent direction of the candidate centerline at that trajectory point and the grid direction. Figure 11 As shown, at trajectory points s1, s2, s3 and s4 respectively, the angles θ1, θ2, θ3 and θ4 between the candidate centerline and the grid direction of the candidate lane grid where each trajectory point is located are obtained.

[0130] Step S1004: Obtain the weight value of each included angle, and perform weighted summation on multiple included angles based on the weight value of each included angle to obtain the road trend cost of the candidate centerline.

[0131] In this embodiment, a weighted summation of multiple included angles can be performed to obtain the road trend value of the candidate centerline. Specifically, the weight value of each included angle can be obtained, and a weighted summation of multiple included angles can be performed based on the weight value of each included angle. As described in the previous example, a weighted summation of multiple included angles θ1, θ2, θ3, θ4 is performed based on the weight values ​​n1, n2, n3, n4 of each included angle to obtain the road trend value A4 of the candidate centerline = n1*θ1 + n2*θ2 + n3*θ3 + n4*θ4. For example, the weight values ​​of the included angles corresponding to each trajectory point can be the same.

[0132] In this embodiment, the road trend of the candidate centerline is evaluated by calculating the angle between the candidate centerline and the grid direction of the lane grid at each trajectory point, and then performing a weighted summation of multiple angles. This provides a prerequisite for subsequently selecting candidate centerlines with lower road trend costs. Therefore, candidate centerlines with smaller steering angles and better alignment with the current road trend can be selected, further improving driving comfort.

[0133] Figure 12 This is a flowchart illustrating an eighth embodiment of a lane selection method for intersections provided in this application. See also... Figure 12 The above step S902 specifically includes the following steps:

[0134] Step S1201: Based on the historical driving trajectory and current driving information of vehicles adjacent to the vehicle, generate a predicted lane for the vehicle adjacent to the vehicle, perform rasterization processing on the predicted lane, and obtain the raster orientation of each predicted lane raster.

[0135] In this embodiment, a predicted lane for adjacent vehicles can be generated using an iterative linear quadratic regulator (iLQR) based on the historical driving trajectories and current driving information of adjacent vehicles. Adjacent vehicles may include those traveling in adjacent lanes; historical driving trajectories may include the driving trajectory information of adjacent vehicles at the previous moment; and current driving information may include the current driving direction information of adjacent vehicles. Figure 13 A schematic diagram illustrating the traffic cost values ​​for determining candidate centerlines provided in this application. (See diagram below.) Figure 13 As shown, the lane lines of the predicted lanes generated based on the historical driving trajectories and current driving information of adjacent vehicles are represented by dashed lines. The two candidate centerlines for vehicles crossing the intersection are represented by solid lines. The predicted lanes are rasterized, and the grid orientation of each predicted lane grid is obtained, as shown by the arrows. For example, the grid orientation can be determined based on the lane line orientation of the predicted lanes.

[0136] Step S1202: Select multiple trajectory points on the candidate center line according to the preset step size.

[0137] In this embodiment, multiple trajectory points can be selected on the candidate centerline according to a preset step size. For example, the preset step size can be 2 meters. Figure 13 As shown, multiple trajectory points u1, u2, u3 and u4 can be selected on the candidate center line according to a preset step size.

[0138] Step S1203: At each trajectory point, obtain the angle between the candidate centerline and the grid direction of the predicted lane grid corresponding to the trajectory point.

[0139] In this embodiment, at each trajectory point, the angle between the candidate centerline and the grid direction of the predicted lane grid corresponding to that trajectory point can be obtained. This angle can be the angle between the tangent direction of the candidate centerline at that trajectory point and the grid direction. Figure 13As shown, since trajectory points u1 and u2 do not fall within the predicted lane grid, the angles between the candidate centerline and the grid direction of the predicted lane grid adjacent to trajectory points u1 and u2 can be obtained, respectively. Since trajectory points u3 and u4 fall within the predicted lane grid, the angles between the candidate centerline and the grid direction of the predicted lane grid containing trajectory points u3 and u4 can be obtained, respectively. For example, the grid direction of the predicted lane grid corresponding to a trajectory point can be translated to the trajectory point to obtain the angle between the tangent direction of the candidate centerline at that trajectory point and the grid direction. The angles β1, β2, β3, and β4 between the candidate centerline and the grid direction of the predicted lane grid corresponding to each trajectory point are obtained at trajectory points u1, u2, u3, and u4, respectively.

[0140] Step S1204: Obtain the weight value of each included angle, and perform weighted summation on multiple included angles based on the weight value of each included angle to obtain the traffic flow cost of the candidate centerline.

[0141] In this embodiment, a weighted summation of multiple included angles can be performed to obtain the traffic flow cost of the candidate centerline. Specifically, the weight value of each included angle can be obtained, and a weighted summation of multiple included angles can be performed based on the weight value of each included angle. As described in the previous example, a weighted summation of multiple included angles β1, β2, β3, β4 is performed based on the weight values ​​w1, w2, w3, w4 of each included angle to obtain the traffic flow cost A5 of the candidate centerline = w1*β1 + w2*β2 + w3*β3 + w4*β4. Exemplarily, the weight values ​​of the included angles corresponding to each trajectory point can be the same.

[0142] In this embodiment, the traffic flow of the candidate centerline is evaluated by calculating the angle between the candidate centerline and the grid direction of the lane grid at each trajectory point, and then performing a weighted summation of multiple angles. This provides a prerequisite for subsequently selecting candidate centerlines with lower traffic flow costs. Therefore, candidate centerlines that do not encroach on the same lane as other vehicles can be selected, further improving driving safety.

[0143] Figure 14 This is a flowchart illustrating Embodiment Nine of the intersection lane selection method provided in this application. See also... Figure 14 The above step S902 specifically includes the following steps:

[0144] Step S1401: Based on the vehicle's navigation information, determine the number of navigation lane changes required for the vehicle to navigate to the candidate lane corresponding to the candidate centerline and the remaining driving distance.

[0145] In this embodiment, the required number of navigation lane changes and the remaining drivable distance for the vehicle to navigate to the candidate lane corresponding to the candidate centerline can be determined based on the vehicle's navigation information. The vehicle's navigation information may include lane change information from the current lane to each candidate lane and the vehicle's distance from the intersection. The number of navigation lane changes represents the number of lane changes required for the vehicle to move from the current lane to a candidate lane; the remaining drivable distance represents the remaining distance the vehicle can travel to cross the intersection. For example, if a vehicle needs to turn left to cross an intersection, and the vehicle is currently traveling in the straight lane, the remaining drivable distance is the distance from the vehicle's current position to the current intersection. If the vehicle does not change lanes to the left-turn lane within this distance, it will be unable to cross the intersection according to the navigation. If the vehicle is currently traveling in the left-turn lane, the remaining drivable distance is greater than the distance from the vehicle's current position to the current intersection (e.g., the distance to the next intersection or the distance to the navigation endpoint).

[0146] Step S1402: Obtain the weight value corresponding to the number of navigation lane changes and the weight value corresponding to the remaining drivable distance.

[0147] In this embodiment, the weight value corresponding to the number of navigation lane changes and the weight value corresponding to the remaining drivable distance can be obtained separately. For example, the weight value corresponding to the number of navigation lane changes can be a number greater than zero; the weight value corresponding to the remaining drivable distance can be a number less than zero.

[0148] Step S1403: Based on the weight value corresponding to the number of navigation lane changes and the weight value corresponding to the remaining drivable distance, perform a weighted summation of the number of navigation lane changes and the remaining drivable distance to obtain the navigation cost of the candidate centerline.

[0149] In this embodiment, the navigation cost of the candidate centerline can be obtained by weighted summation of the number of navigation lane changes and the remaining drivable distance. As described in the previous example, the navigation cost of the candidate centerline A6 = b1*c + b2*l is obtained by weighted summation of the number of navigation lane changes c and the remaining drivable distance l based on the weight value b1 corresponding to the number of navigation lane changes and the weight value b2 corresponding to the remaining drivable distance.

[0150] In this embodiment, the navigation cost of the candidate centerlines is evaluated by obtaining the number of navigation lane changes required and the remaining drivable distance for each candidate centerline in the corresponding lane, and then performing a weighted summation of these values. This provides a prerequisite for subsequently selecting candidate centerlines with lower navigation costs. Therefore, candidate centerlines with fewer lane changes and longer remaining drivable distances can be selected, further improving driving comfort.

[0151] Figure 15This is a flowchart illustrating Embodiment 10 of the intersection lane selection method provided in this application. See also... Figure 15 The above step S902 specifically includes the following steps:

[0152] Step S1501: Determine the vehicle speed in the candidate lane corresponding to the candidate centerline based on the traffic speed information of multiple candidate lanes.

[0153] In this embodiment, the traffic flow speed information of the candidate lane may include the average speed of vehicles traveling in the candidate lane, or it may include historical traffic flow images of the candidate lane obtained based on visual perception technology, and the predicted traffic flow speed of the candidate lane obtained.

[0154] Step S1502: Determine the driving efficiency value of the candidate centerline based on the traffic flow speed.

[0155] In this embodiment, the driving efficiency value of the candidate centerline can be determined based on the traffic flow speed. For example, a correspondence between traffic flow speed and driving efficiency value A7 can be set, or a proportional coefficient can be set so that the traffic flow speed and driving efficiency value are inversely proportional, that is, the higher the traffic flow speed, the smoother the lane, and the lower the driving efficiency value.

[0156] In this embodiment, the driving efficiency of the candidate centerline is evaluated by the traffic flow speed in the candidate lane corresponding to the candidate centerline, providing a prerequisite for subsequently selecting candidate centerlines with lower driving efficiency costs. Therefore, candidate centerlines corresponding to candidate lanes with higher traffic flow speeds can be selected, further improving driving comfort.

[0157] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0158] Figure 16 This is a schematic diagram illustrating the structure of an embodiment of a lane selection processing device for intersections provided in this application. Figure 16As shown, the lane selection processing device 160 at the intersection includes an acquisition module 161 and a processing module 162. The acquisition module 161 acquires vehicle information and lane information of multiple candidate lanes at preset time intervals. The vehicle information includes the vehicle's current location and historical centerline information, and the lane information includes the position information and obstacle information of the candidate lane. The processing module 162 generates a candidate centerline for each candidate lane based on the vehicle's current location and the lane's position information. The processing module 162 also determines, for each candidate centerline, a consistency value, a comfort value, and a safety value based on the vehicle's current location, historical centerline information, and obstacle information of the candidate lane, and determines the value of the candidate centerline based on these values. The processing module 162 further determines the minimum value among the multiple candidate centerlines and identifies the candidate centerline corresponding to the minimum value as the vehicle's centerline, controlling the vehicle to travel along that centerline.

[0159] In one possible implementation, the processing module 162 is specifically used to select multiple trajectory points on the candidate centerline and the historical centerline according to a preset step size; wherein, the trajectory points on the candidate centerline correspond one-to-one with the trajectory points on the historical centerline; for each trajectory point on the candidate centerline, calculate the distance between the trajectory point and the corresponding trajectory point on the historical centerline; obtain the weight value of each distance, and perform weighted summation of multiple distances according to the weight value of each distance to obtain the consistency cost of the candidate centerline.

[0160] In one possible implementation, the processing module 162 is specifically used to generate the vehicle's driving trajectory based on the vehicle's current location information and the candidate centerline; select multiple trajectory points on the driving trajectory according to a preset step size, and obtain the speed change information of each trajectory point; obtain the weight value of each speed change information, and perform weighted summation processing on the speed change information of multiple trajectory points according to the weight value of each speed change information to obtain the comfort value of the candidate centerline.

[0161] In one possible implementation, the processing module 162 is specifically configured to generate the vehicle's driving trajectory based on the vehicle's current location information and the candidate centerline; obtain the minimum distance between each obstacle on the candidate lane and the driving trajectory, and determine the minimum value of multiple minimum distances; and determine the safety cost of the candidate centerline based on the minimum value.

[0162] In one possible implementation, the acquisition module 161 is further configured to acquire the consistency cost threshold, the comfort cost threshold, and the safety cost threshold of the consistency cost value, respectively; the processing module 162 is specifically configured to determine the cost value of the candidate centerline based on the consistency cost value, the comfort cost value, and the safety cost value when it is determined that the consistency cost value is less than the consistency cost threshold, the comfort cost value is less than the comfort cost threshold, and the safety cost value is less than the safety cost threshold.

[0163] In one possible implementation, the acquisition module 161 is further configured to acquire at least one of the following: navigation information of the vehicle, lane line information and traffic flow speed information of multiple candidate lanes, and historical driving trajectory and current driving information of vehicles adjacent to the vehicle; the processing module 162 is further configured to determine at least one of the following: road trend value, traffic flow value, navigation value, and driving efficiency value of the candidate centerline, based on the navigation information of the vehicle, lane line information and traffic flow speed information of multiple candidate lanes, and historical driving trajectory and current driving information of vehicles adjacent to the vehicle; specifically, the processing module 162 is configured to determine the value of the candidate centerline based on the consistency value, comfort value, safety value, and at least one of the following: road trend value, traffic flow value, navigation value, and driving efficiency value.

[0164] In one possible implementation, the processing module 162 is specifically used to perform rasterization processing on the candidate lane corresponding to the candidate centerline based on the lane line information of multiple candidate lanes, and obtain the grid direction of each candidate lane grid; select multiple trajectory points on the candidate centerline according to a preset step size; at each trajectory point, obtain the angle between the candidate centerline and the grid direction of the candidate lane grid where the trajectory point is located; obtain the weight value of each angle, and perform weighted summation processing on multiple angles according to the weight value of each angle to obtain the road trend value of the candidate centerline.

[0165] In one possible implementation, the processing module 162 is specifically configured to generate a predicted lane for vehicles adjacent to the vehicle based on the historical driving trajectory and current driving information of vehicles adjacent to the vehicle; perform rasterization processing on the predicted lane and obtain the raster orientation of each predicted lane raster; select multiple trajectory points on the candidate centerline according to a preset step size; at each trajectory point, obtain the angle between the candidate centerline and the raster orientation of the predicted lane raster corresponding to the trajectory point; obtain the weight value of each angle; and perform weighted summation processing on multiple angles based on the weight value of each angle to obtain the traffic flow cost of the candidate centerline.

[0166] In one possible implementation, the processing module 162 is specifically configured to determine, based on the vehicle's navigation information, the number of navigation lane changes required by the vehicle in the candidate lane corresponding to the candidate centerline and the remaining drivable distance; obtain the weight values ​​corresponding to the number of navigation lane changes and the remaining drivable distance respectively; and perform weighted summation on the number of navigation lane changes and the remaining drivable distance based on the weight values ​​corresponding to the number of navigation lane changes and the remaining drivable distance to obtain the navigation cost of the candidate centerline.

[0167] In one possible implementation, the processing module 162 is specifically used to determine the vehicle's traffic speed in the candidate lane corresponding to the candidate centerline based on the traffic speed information of multiple candidate lanes; and to determine the driving efficiency value of the candidate centerline based on the traffic speed.

[0168] The intersection lane selection processing device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0169] Figure 17 This is a schematic diagram illustrating the structure of an embodiment of an electronic device provided in this application. Figure 17 As shown, the electronic device 170 includes: a processor 171, a memory 172, and a communication interface 173; wherein, the memory 172 is used to store executable instructions of the processor 171; the processor 171 is configured to execute the technical solutions of any of the foregoing method embodiments by executing the executable instructions.

[0170] Optionally, the memory 172 can be either standalone or integrated with the processor 171.

[0171] Optionally, when the memory 172 is a device independent of the processor 171, the electronic device 170 may further include a bus 174 for connecting the aforementioned devices.

[0172] The electronic device is used to execute the technical solutions in any of the foregoing method embodiments. Its implementation principle and technical effect are similar, and will not be described again here.

[0173] This application also provides a vehicle that includes electronic devices as provided in any of the foregoing embodiments.

[0174] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the technical solutions provided in any of the foregoing embodiments.

[0175] This application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the technical solutions provided in any of the foregoing embodiments.

[0176] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0177] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for lane selection at intersections, characterized in that, include: At preset intervals, vehicle information and lane information of multiple candidate lanes are acquired; wherein, the vehicle information includes the vehicle's current location information and historical centerline information, and the lane information includes the location information and obstacle information of the candidate lanes; For each candidate lane, a candidate centerline is generated based on the current location information of the vehicle and the location information of the candidate lane. For each candidate centerline, based on the vehicle's current location information, historical centerline information, and obstacle information of the candidate lane, the consistency value, comfort value, and safety value of the candidate centerline are determined respectively, and the value of the candidate centerline is determined based on the consistency value, comfort value, and safety value. The minimum value among the multiple candidate centerlines is determined, and the candidate centerline corresponding to the minimum value is determined as the centerline of the vehicle, and the vehicle is controlled to travel along the centerline.

2. The method according to claim 1, characterized in that, Based on the vehicle's historical centerline information, the consistency value of the candidate centerline is determined, including: According to a preset step size, multiple trajectory points are selected on the candidate center line and the historical center line respectively; wherein, the trajectory points on the candidate center line correspond one-to-one with the trajectory points on the historical center line. For each trajectory point on the candidate center line, calculate the distance between the trajectory point and the corresponding trajectory point on the historical center line; Obtain the weight value of each distance, and perform a weighted summation of multiple distances based on the weight value of each distance to obtain the consistency cost of the candidate centerline.

3. The method according to claim 1, characterized in that, Based on the vehicle's current location information, the comfort value of the candidate centerline is determined, including: The vehicle's driving trajectory is generated based on the vehicle's current location information and the candidate centerline; According to the preset step size, select multiple trajectory points on the driving trajectory and obtain the speed change information of each trajectory point; Obtain the weight value of each velocity change information, and perform weighted summation on the velocity change information of multiple trajectory points based on the weight value of each velocity change information to obtain the comfort cost of the candidate centerline.

4. The method according to claim 1, characterized in that, Based on the vehicle's current location information and the obstacle information of the lane to be selected, the safety cost of the candidate centerline is determined, including: The vehicle's driving trajectory is generated based on the vehicle's current location information and the candidate centerline; Obtain the minimum distance between each obstacle on the candidate lane corresponding to the candidate centerline and the driving trajectory, and determine the minimum value of multiple minimum distances; The safety cost of the candidate centerline is determined based on the minimum value.

5. The method according to any one of claims 1 to 4, characterized in that, Before determining the value of the candidate centerline based on the consistency value, comfort value, and safety value, the method further includes: Obtain the consistency cost threshold, comfort cost threshold, and safety cost threshold of the consistency cost value, respectively; The value of the candidate centerline is determined based on the consistency value, comfort value, and safety value, including: When it is determined that the consistency cost value is less than the consistency cost threshold, the comfort cost value is less than the comfort cost threshold, and the safety cost value is less than the safety cost threshold, the cost value of the candidate centerline is determined based on the consistency cost value, the comfort cost value, and the safety cost value.

6. The method according to any one of claims 1 to 4, characterized in that, The method further includes: Obtain at least one of the following: navigation information of the vehicle, lane line information and traffic speed information of multiple candidate lanes, and historical driving trajectory and current driving information of vehicles adjacent to the vehicle. Based on the vehicle's navigation information, lane line information and traffic flow speed information of multiple candidate lanes, and at least one of the historical driving trajectory and current driving information of vehicles adjacent to the vehicle, determine at least one of the road trend value, traffic flow value, navigation value and driving efficiency value of the candidate centerline. The step of determining the cost value of the candidate centerline based on the consistency cost value, comfort cost value, and safety cost value includes: The value of the candidate centerline is determined based on at least one of the following: consistency value, comfort value, safety value, road trend value, traffic flow value, navigation value, and driving efficiency value.

7. The method according to claim 6, characterized in that, Based on lane line information from multiple candidate lanes, the road trend value of the candidate centerline is determined, including: Based on the lane line information of multiple candidate lanes, the candidate lanes corresponding to the candidate center lines are rasterized, and the raster direction of each candidate lane raster is obtained. According to the preset step size, select multiple trajectory points on the candidate center line; At each trajectory point, the angle between the candidate centerline and the grid direction of the candidate lane grid where the trajectory point is located is obtained; Obtain the weight value of each of the included angles, and perform a weighted summation of multiple included angles based on the weight value of each of the included angles to obtain the road trend cost value of the candidate centerline; And / or, Based on the historical driving trajectories and current driving information of vehicles adjacent to the vehicle, the traffic flow value of the candidate centerline is determined, including: Based on the historical driving trajectories and current driving information of vehicles adjacent to the vehicle, a predicted lane for the vehicle adjacent to the vehicle is generated, the predicted lane is rasterized, and the raster orientation of each predicted lane raster is obtained. According to the preset step size, select multiple trajectory points on the candidate center line; At each trajectory point, the angle between the candidate centerline and the grid direction of the predicted lane grid corresponding to the trajectory point is obtained; Obtain the weight value of each of the included angles, and perform a weighted summation of multiple included angles based on the weight value of each of the included angles to obtain the traffic flow cost of the candidate centerline.

8. The method according to claim 6, characterized in that, Based on the vehicle's navigation information, the navigation value of the candidate centerline is determined, including: Based on the vehicle's navigation information, determine the number of navigation lane changes required for the vehicle in the candidate lane corresponding to the candidate centerline and the remaining drivable distance; Obtain the weight values ​​corresponding to the number of navigation lane changes and the weight values ​​corresponding to the remaining drivable distance, respectively. Based on the weight values ​​corresponding to the number of navigation lane changes and the weight values ​​corresponding to the remaining drivable distance, the number of navigation lane changes and the remaining drivable distance are weighted and summed to obtain the navigation cost of the candidate centerline; And / or, Based on the traffic speed information of multiple candidate lanes, the driving efficiency cost of the candidate centerline is determined, including: Based on the traffic flow speed information of multiple candidate lanes, determine the traffic flow speed of the vehicle in the candidate lane corresponding to the candidate centerline; The driving efficiency value of the candidate centerline is determined based on the traffic flow speed.

9. An electronic device, characterized in that, include: Processor, memory, communication interface; The memory is used to store the executable instructions of the processor; The processor is configured to execute the intersection selection processing method of any one of claims 1 to 8 by executing the executable instructions.

10. A car, characterized in that, include: The electronic device as described in claim 9.

Images