Path selection method, device and electronic equipment for autonomous vehicle

Abstract

The application discloses a path selection method and device of an automatic driving vehicle and electronic equipment, and the method comprises the following steps: acquiring current positioning data of the automatic driving vehicle and corresponding path planning results, wherein the path planning results comprise a plurality of alternative lane-level paths; acquiring predicted positioning data of the automatic driving vehicle and obstacles based on the current positioning data of the automatic driving vehicle; determining driving costs of the alternative lane-level paths according to the predicted positioning data of the automatic driving vehicle and the obstacles, wherein the driving cost is used for representing congestion degree of the alternative lane-level path; and determining a target lane-level path according to the driving costs of the alternative lane-level paths. The application considers the influence of traffic flow and pedestrian flow conditions on the driving efficiency of the ego vehicle in a future period of time, evaluates the spatial congestion conditions of the alternative lane-level paths in the future period of time in combination with the predicted positioning data of the ego vehicle and the obstacles, provides a reference for the driving decision of the ego vehicle, and improves the driving efficiency.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a path selection method, apparatus and electronic device for autonomous vehicles. Background Technology

[0002] Advanced Driver Assistance Systems (ADAS) utilize various onboard sensors and decision-making and control algorithms to enable vehicles to perform behaviors such as lane changing, merging, overtaking, and following. Due to the complexity of real-world road scenarios, autonomous vehicles need to consider the actual traffic and pedestrian flow while fulfilling their target driving tasks. Through comprehensive decision-making algorithms, they select lanes to complete vehicle behaviors such as lane changing and merging.

[0003] Currently, when making lane decisions, existing solutions only consider the current traffic and pedestrian flow information and do not take into account the future traffic and pedestrian flow conditions. Therefore, the decisions made can only reflect the current situation and cannot adapt to future traffic and pedestrian flow conditions, resulting in low overall driving efficiency of autonomous vehicles. Summary of the Invention

[0004] This application provides a path selection method, apparatus, and electronic device for autonomous vehicles to improve driving efficiency by evaluating the driving costs of autonomous vehicles.

[0005] The embodiments of this application adopt the following technical solutions:

[0006] In a first aspect, embodiments of this application provide a path selection method for an autonomous vehicle, wherein the method includes:

[0007] The current location data of the autonomous vehicle and the path planning results corresponding to the current location data of the autonomous vehicle are obtained, and the path planning results include multiple alternative lane-level paths.

[0008] Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle and the predicted location data of the obstacles are obtained.

[0009] Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the driving cost of each alternative lane-level path is determined, and the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0010] Based on the driving cost of each alternative lane-level path, the target lane-level path is determined from multiple alternative lane-level paths.

[0011] Optionally, obtaining the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles based on the current positioning data of the autonomous vehicle includes:

[0012] Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle within a preset time period is determined.

[0013] Based on the current location data of the autonomous vehicle, corresponding obstacle perception data is obtained, and the predicted location data of the obstacle within a preset time period is predicted based on the obstacle perception data.

[0014] Optionally, determining the predicted location data of the autonomous vehicle within a preset time period based on the current location data of the autonomous vehicle includes:

[0015] Establish the Frenet coordinate system corresponding to the candidate lane-level path based on the target lane in the candidate lane-level path;

[0016] The current location data of the autonomous vehicle is converted to the Frenet coordinate system corresponding to the alternative lane-level path;

[0017] Based on the current positioning data of the autonomous vehicle in the Frenet coordinate system corresponding to the alternative lane-level path, the predicted positioning data of the autonomous vehicle within a preset time period is determined.

[0018] Optionally, the step of obtaining corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacle within a preset time period based on the obstacle perception data includes:

[0019] Establish the Frenet coordinate system corresponding to the candidate lane-level path based on the target lane in the candidate lane-level path;

[0020] Based on the current location data of the autonomous vehicle, the corresponding obstacle perception data is obtained, and the predicted location data of the obstacle within a preset time period is predicted based on the obstacle perception data.

[0021] The predicted location data of the obstacles within the preset time period are converted to the Frenet coordinate system corresponding to the alternative lane-level path to obtain the predicted location data of the obstacles corresponding to the alternative lane-level path.

[0022] Optionally, converting the predicted location data of the obstacle within the preset time period to the Frenet coordinate system corresponding to the candidate lane-level path to obtain the predicted location data of the obstacle corresponding to the candidate lane-level path includes:

[0023] Determine the lane width of the target lane in the candidate lane-level paths;

[0024] Based on the lane width of the target lane in the candidate lane-level path, the predicted location data of obstacles in the Frenet coordinate system corresponding to the candidate lane-level path are filtered to obtain the predicted location data of obstacles corresponding to the candidate lane-level path.

[0025] Optionally, determining the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles includes:

[0026] Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the longitudinal distance between the autonomous vehicle and the obstacle is determined;

[0027] The driving cost corresponding to each of the alternative lane-level paths is determined based on the longitudinal distance between the autonomous vehicle and the obstacle.

[0028] Optionally, both the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle include predicted positions at multiple prediction times. Determining the travel cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle includes:

[0029] Based on the predicted positions of the autonomous vehicle and the obstacle at each predicted time, the longitudinal distance between the autonomous vehicle and the obstacle at each predicted time is determined.

[0030] Based on the longitudinal distance between the autonomous vehicle and the obstacle at each time point, the driving cost of each alternative lane-level path is determined using a preset weighted summation algorithm.

[0031] Secondly, embodiments of this application also provide a route selection device for an autonomous vehicle, wherein the device includes:

[0032] The first acquisition unit is used to acquire the current positioning data of the autonomous vehicle and the path planning result corresponding to the current positioning data of the autonomous vehicle, wherein the path planning result includes multiple alternative lane-level paths.

[0033] The second acquisition unit is used to acquire the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles based on the current positioning data of the autonomous vehicle.

[0034] The first determining unit is used to determine the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, wherein the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0035] The second determining unit is used to determine the target lane-level path from multiple candidate lane-level paths based on the travel cost of each candidate lane-level path.

[0036] Thirdly, embodiments of this application also provide an electronic device, including:

[0037] Processor; and

[0038] A memory is configured to store computer-executable instructions, which, when executed, cause the processor to perform any of the aforementioned path selection methods for autonomous vehicles.

[0039] Fourthly, embodiments of this application also provide a computer-readable storage medium storing one or more programs that, when executed by an electronic device including multiple applications, cause the electronic device to perform any of the aforementioned path selection methods for autonomous vehicles.

[0040] The above-mentioned at least one technical solution adopted in the embodiments of this application can achieve the following beneficial effects: The path selection method for autonomous vehicles in the embodiments of this application first obtains the current positioning data of the autonomous vehicle and the path planning result corresponding to the current positioning data of the autonomous vehicle, the path planning result including multiple candidate lane-level paths; then, based on the current positioning data of the autonomous vehicle, it obtains the predicted positioning data of the autonomous vehicle and the predicted positioning data of obstacles; then, based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of obstacles, it determines the driving cost of each candidate lane-level path, the driving cost is used to characterize the congestion degree of each candidate lane-level path; finally, based on the driving cost of each candidate lane-level path, it determines the target lane-level path among multiple candidate lane-level paths. The path selection method for autonomous vehicles in the embodiments of this application considers the impact of traffic flow and pedestrian flow on the driving efficiency of the autonomous vehicle in the future, and evaluates the spatial congestion of each candidate lane-level path in the future by combining the predicted positioning data of the autonomous vehicle and obstacles, providing a reference for the driving decision of the autonomous vehicle and improving the driving efficiency of the autonomous vehicle. Attached Figure Description

[0041] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0042] Figure 1 This is a flowchart illustrating a path selection method for an autonomous vehicle according to an embodiment of this application.

[0043] Figure 2 This is a schematic diagram of the structure of a path selection device for an autonomous vehicle according to an embodiment of this application;

[0044] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0046] The technical solutions provided by the various embodiments of this application are described in detail below with reference to the accompanying drawings.

[0047] This application provides a path selection method for autonomous vehicles, such as... Figure 1 The diagram provided illustrates a path selection method for an autonomous vehicle according to an embodiment of this application. The method includes at least the following steps S110 to S140:

[0048] Step S110: Obtain the current location data of the autonomous vehicle and the path planning result corresponding to the current location data of the autonomous vehicle. The path planning result includes multiple alternative lane-level paths.

[0049] In this embodiment of the application, when selecting the driving path of an autonomous vehicle (hereinafter referred to as "vehicle"), it is necessary to first obtain the current positioning data of the vehicle and the corresponding path planning results. The current positioning data can be obtained based on the positioning device on the vehicle, such as an inertial navigation positioning device, and may specifically include information such as the current position, speed and acceleration of the vehicle.

[0050] The path planning result corresponding to the current positioning data can be obtained through the path planning module on the vehicle. The path planning module can provide multiple alternative lane-level paths based on existing lane-level path planning algorithms. Of course, those skilled in the art can flexibly choose how to perform path planning according to actual needs, and no specific limitations are made here.

[0051] Step S120: Based on the current positioning data of the autonomous vehicle, obtain the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles.

[0052] Based on the current location data of the vehicle, it is also necessary to further obtain the predicted location data of the vehicle and the predicted location data of obstacles in the future. Here, the predicted location data of the vehicle and the predicted location data of obstacles can be obtained by predicting each alternative lane-level path. Obstacles can refer to other vehicles on the road besides the vehicle, as well as pedestrians and other dynamic targets. The predicted location data of obstacles can be obtained through the obstacle perception and prediction module on the vehicle. The specific method of acquisition can be flexibly selected by those skilled in the art according to actual needs, and no specific limitation is made here.

[0053] Step S130: Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, determine the driving cost of each alternative lane-level path, wherein the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0054] By comparing the predicted location data of the vehicle obtained in the aforementioned steps with the predicted location data of the corresponding obstacles, the driving cost of each alternative lane-level path can be calculated. The driving cost here is mainly used to characterize the spatial congestion of each alternative lane-level path in the future. The higher the driving cost of the alternative lane-level path, the higher the congestion level of the alternative lane-level path in the future, and vice versa.

[0055] Step S140: Determine the target lane-level path from among the multiple candidate lane-level paths based on the travel cost of each candidate lane-level path.

[0056] Different driving costs reflect the varying degrees of congestion on each alternative lane-level path over a future period. Therefore, when making driving decisions, the vehicle can choose the alternative lane-level path with the lowest driving cost, i.e., the least congestion, as the final lane-level path, thereby ensuring the vehicle's driving efficiency.

[0057] The path selection method for autonomous vehicles in this application takes into account the impact of traffic flow and pedestrian flow on the driving efficiency of the autonomous vehicle in the future. By combining the predicted positioning data of the autonomous vehicle and obstacles, the spatial congestion of each alternative lane-level path in the future is evaluated, which provides a reference for the driving decision of the autonomous vehicle and improves the driving efficiency of the autonomous vehicle.

[0058] In some embodiments of this application, the path planning result corresponding to the current positioning data of the autonomous vehicle is obtained in the following manner: obtaining the reference trajectory of the autonomous vehicle and the local high-precision map data corresponding to the autonomous vehicle, wherein the reference trajectory includes the termination trajectory point of the reference trajectory; determining the target lane set of the autonomous vehicle based on the termination trajectory point of the reference trajectory and the local high-precision map data corresponding to the autonomous vehicle; generating the alternative path set of the vehicle based on the target lane set of the vehicle and a preset map algorithm, wherein the alternative path set includes multiple alternative lane-level paths.

[0059] First, in generating the alternative path set S-lane for the vehicle, the embodiments of this application can first construct a directed graph G and a corresponding matrix M using local high-precision map data. The directed graph and matrix describe the connection relationship between all lanes within the area corresponding to the local high-precision map, which can greatly improve the efficiency of path planning.

[0060] Then, initialize the dynamic target lane set S-lane with an initial length of 0, and take the last point Pn of the reference trajectory Traj, handling the following cases:

[0061] 1) If Pn is the destination, then take the lane Lane0 where Pn is located and insert it into the target lane set S-lane;

[0062] 2) If Pn is not the destination and the lane where Pn is located is Lane 0, then search for all other lanes of the same type LaneN in the same road segment of Lane 0 in the local high-precision map according to the lane type of Lane 0 (such as straight, left turn, right turn, straight + turn) and save them in S-lane.

[0063] 3) If Pn is not the destination and Pn is located on the lane connection line, then search for all other lanes of the same type LaneN in the same road segment of Lane0 in the local high-precision map according to the lane type of the target lane Lane0 connected by the connection line, and save them in S-lane.

[0064] Finally, determine the lane-ego where the vehicle is currently located, and extract the target lane from the S-lane in sequence. Calculate all connected paths based on the directed graph G to obtain the set of candidate paths, Routs. Here, the depth-first search (DFS) algorithm can be used to traverse the graph G.

[0065] Furthermore, it should be noted that the multiple alternative lane-level paths finally obtained in the embodiments of this application can be regarded as the results of preliminary screening. The preliminary screening strategy here may include filtering loop paths in each alternative lane-level path, assessing the collision risk between the vehicle and obstacles in each alternative lane-level path, etc.

[0066] This application embodiment, based on reference trajectory and high-precision map data, constructs a directed graph and then uses a connectivity search algorithm to generate multiple alternative lane-level paths that comply with traffic regulations and other constraints. This provides a prerequisite for subsequent scenario decisions and avoids the problem that some road sections cannot reach the destination due to vehicle lane-changing behavior.

[0067] In some embodiments of this application, obtaining the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle based on the current positioning data of the autonomous vehicle includes: determining the predicted positioning data of the autonomous vehicle within a preset time period based on the current positioning data of the autonomous vehicle; obtaining the corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacle within the preset time period based on the obstacle perception data.

[0068] This application embodiment can determine the vehicle's predicted location data for a future period of time using the vehicle's current location data. Since each alternative lane-level path is planned for the vehicle's future driving path, when determining the vehicle's predicted location data for a future period of time, points can be collected on each alternative lane-level path at certain time intervals s to obtain the vehicle's predicted location data sequence E = {e}. s1 e s2 e s3 ...e sn}

[0069] In this embodiment of the application, when acquiring the predicted location data of obstacles, obstacle perception data can be acquired first. This obstacle perception data can be obtained by sensing and identifying environmental information around the vehicle using sensors such as vision sensors, millimeter-wave radar, and lidar. Based on the historical trajectory and current state of each obstacle, a predicted trajectory can be generated for each obstacle, serving as the predicted location data. The historical trajectory of an obstacle can include its historical position and velocity; the current state of an obstacle can include its current position, velocity, acceleration, and orientation; and the predicted trajectory of an obstacle can include its predicted position and velocity.

[0070] It should be noted that, since the frequency of the obstacle prediction and positioning data output by the obstacle perception and prediction module may differ from the sampling frequency of the vehicle's prediction and positioning data, the obstacle prediction and positioning data can be sampled at the same time intervals based on the sampling frequency of the vehicle's prediction and positioning data. This time synchronization process generates an obstacle prediction and positioning data sequence P = {p...} at the same time as the vehicle's prediction and positioning data sequence E. s1 p s2 p s3 ...p sn When there are multiple obstacles, the predicted localization data sequence P corresponding to each obstacle is obtained through separate processing.

[0071] In some embodiments of this application, determining the predicted positioning data of the autonomous vehicle within a preset time period based on the current positioning data of the autonomous vehicle includes: establishing a Frenet coordinate system corresponding to the alternative lane-level path based on the target lane in the alternative lane-level path; converting the current positioning data of the autonomous vehicle to the Frenet coordinate system corresponding to the alternative lane-level path; and determining the predicted positioning data of the autonomous vehicle within the preset time period based on the current positioning data of the autonomous vehicle in the Frenet coordinate system corresponding to the alternative lane-level path.

[0072] In determining the predicted positioning data of the vehicle, this application embodiment can establish a Frenet coordinate system corresponding to the target lane in each candidate lane-level path. The Frenet coordinate system is a lane coordinate system that describes the position of the vehicle relative to the road. The vertical axis represents the distance along the road, and the horizontal axis represents the lateral vertical displacement relative to the longitudinal line. When the vehicle moves forward in the Frenet coordinate system, its trajectory in the lane is a straight line, which can greatly simplify the difficulty of trajectory planning.

[0073] Then, the vehicle's current positioning data is projected into the corresponding Frenet coordinate system. Based on the vehicle's current position, speed, acceleration, and other information in the Frenet coordinate system, multiple position information of the vehicle within a certain time period are determined at certain time intervals. Since this embodiment mainly assesses the spatial congestion situation in the future based on the predicted positioning data between the vehicle and obstacles, the longitudinal position of the vehicle can be used for the position information here.

[0074] In some embodiments of this application, the step of obtaining corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacle within a preset time period based on the obstacle perception data includes: establishing a Frenet coordinate system corresponding to the alternative lane-level path based on the target lane in the alternative lane-level path; obtaining corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacle within a preset time period based on the obstacle perception data; and converting the predicted positioning data of the obstacle within the preset time period to the Frenet coordinate system corresponding to the alternative lane-level path to obtain the predicted positioning data of the obstacle corresponding to the alternative lane-level path.

[0075] In this embodiment of the application, when predicting obstacle location data, a Frenet coordinate system corresponding to each candidate lane-level path can be established first. That is, both the predicted positioning data of the vehicle and the predicted positioning data of the obstacle must be converted to the Frenet coordinate system corresponding to each candidate lane-level path for subsequent processing. However, for obstacles, since the candidate lane-level paths are not generated for the future driving trajectory of the obstacle, before converting the obstacle's positioning data to the Frenet coordinate system, the obstacle perception and prediction module can first predict the obstacle's position and speed and other positioning data over a period of time in the future, and then convert the predicted positioning data of the obstacle over a period of time to the Frenet coordinate system.

[0076] In some embodiments of this application, the step of converting the predicted location data of the obstacle within the preset time period to the Frenet coordinate system corresponding to the alternative lane-level path to obtain the predicted location data of the obstacle corresponding to the alternative lane-level path includes: determining the lane width of the target lane in the alternative lane-level path; and filtering the predicted location data of the obstacle in the Frenet coordinate system corresponding to the alternative lane-level path according to the lane width of the target lane in the alternative lane-level path to obtain the predicted location data of the obstacle corresponding to the alternative lane-level path.

[0077] Because the predicted location data of obstacles in the future may not necessarily conflict with the predicted location data of the vehicle or affect the vehicle's driving. For example, the lane that the predicted trajectory of the obstacle will pass through may be different from the target lane in the alternative lane-level path that the vehicle will pass through, that is, they are not in the same lane.

[0078] Based on this, in this embodiment, for each candidate lane-level path, the width of the target lane in the candidate lane-level path can be determined first. Then, the predicted position of the obstacle is compared with the width of the target lane to determine whether the predicted position of the obstacle falls within the target lane, that is, whether the obstacle will enter the target lane that the vehicle may enter. If the predicted position of the obstacle falls within the target lane, it means that if the vehicle chooses the candidate lane-level path corresponding to the target lane, its driving in the future is likely to be affected by the obstacle, which will affect the assessment of the driving cost of the candidate lane-level path. If the predicted position of the obstacle does not fall within the target lane, it means that if the vehicle chooses the candidate lane-level path corresponding to the target lane, its driving in the future will hardly be affected by the obstacle. Therefore, the assessment of the driving cost of the candidate lane-level path does not need to consider the obstacle. This is to filter out the predicted location data of obstacles that may spatially affect the vehicle's driving in the future.

[0079] In some embodiments of this application, determining the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle includes: determining the longitudinal distance between the autonomous vehicle and the obstacle based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle; and determining the driving cost corresponding to each alternative lane-level path based on the longitudinal distance between the autonomous vehicle and the obstacle.

[0080] In determining the driving cost of each alternative lane-level path in this embodiment, the longitudinal distance between the vehicle and the obstacle can be calculated based on the predicted positioning data of the vehicle and the obstacle. Specifically, it can be calculated based on the longitudinal position in the predetermined positioning data of the two. Since the obstacle and the vehicle are located in the same target lane, the distance between the lateral positions is not very meaningful for judging the degree of congestion. However, the distance between the longitudinal positions can better reflect the traffic flow and pedestrian flow in the current target lane. Therefore, by calculating the longitudinal distance between the vehicle and the obstacle, the driving cost corresponding to each alternative lane-level path can be evaluated, that is, the spatial congestion.

[0081] In some embodiments of this application, the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle both include predicted positions at multiple prediction times. The step of determining the driving cost of each candidate lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle includes: determining the longitudinal distance between the autonomous vehicle and the obstacle at each prediction time based on the predicted positions of the autonomous vehicle and the obstacle at each prediction time; and determining the driving cost of each candidate lane-level path using a preset weighted summation algorithm based on the longitudinal distance between the autonomous vehicle and the obstacle at each prediction time.

[0082] The predicted positioning data for both the vehicle and obstacles in this application embodiment includes predicted positions at multiple times. For example, the predicted positioning data for the vehicle can be represented as a sequence E = {e s1 e s2 e s3 ...e sn The predicted location data of obstacles can be represented as a sequence P = {p} s1 p s2 p s3 ...p sn Therefore, when calculating the longitudinal distance d between the vehicle and the obstacle, it can be calculated according to each time step, for example, it can be expressed in the following form:

[0083]

[0084] After obtaining the longitudinal distance d between the vehicle and the obstacle at each time step, the driving cost of the alternative lane-level path at each time step can be calculated in the following form.

[0085]

[0086] Equation (2) shows that the greater the longitudinal distance, the lower the driving cost and the lower the congestion level at that time. When there are multiple obstacles at a given time, the driving cost calculated based on the longitudinal distance between the vehicle and each obstacle can be summed to obtain the driving cost at that time.

[0087] Then, based on the driving cost of the alternative lane-level paths at each time point... It can calculate the final total cost (COST) of each alternative lane-level path. total :

[0088]

[0089] Here, k is the decay factor. As time accumulates, the error in the obstacle prediction and positioning data increases, leading to increased uncertainty in obstacle prediction. Therefore, the driving cost calculated at the corresponding time point... By assigning relatively smaller weights, the reliability of driving cost assessment is improved.

[0090] Of course, it should be noted that the above formulas (2)-(3) are only one example of calculating the driving cost. Those skilled in the art can flexibly define other forms of calculation methods according to actual needs, as long as they can reflect the negative correlation between the longitudinal distance between the vehicle and the obstacle and the driving cost.

[0091] The above embodiments evaluated the driving cost of each alternative lane path from both temporal and spatial dimensions, and measured the congestion status of each alternative lane path, providing a strong reference for the driving decisions of autonomous vehicles.

[0092] In some embodiments of this application, the current positioning data of the autonomous vehicle further includes the current speed and current acceleration. The step of determining the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle includes: determining the current driving state of the autonomous vehicle based on the current speed and the current acceleration; and determining the driving cost of each alternative lane-level path based on the current driving state of the autonomous vehicle, the predicted positioning data of the autonomous vehicle, and the predicted positioning data of the obstacle.

[0093] In determining the travel cost of each candidate lane-level path in this embodiment, the current driving state of the vehicle can be determined first, such as whether it is stationary or moving. Specifically, this can be determined based on the vehicle's current speed and current acceleration. If both the vehicle's current speed and current acceleration are 0, it means the vehicle is stationary. In this case, position prediction is not required, and the travel cost (COST) of each candidate lane-level path can be calculated. total Setting both values ​​to 0 means that when the vehicle is stationary, no travel cost calculation is needed. However, if the vehicle's current speed and acceleration are not zero, it indicates that the vehicle is in motion, and in this case, position prediction and travel cost calculation are required.

[0094] This application also provides a path selection device 200 for an autonomous vehicle, such as... Figure 2 The diagram shows a schematic representation of a path selection device for an autonomous vehicle according to an embodiment of this application. The device 200 includes: a first acquisition unit 210, a second acquisition unit 220, a first determination unit 230, and a second determination unit 240, wherein:

[0095] The first acquisition unit 210 is used to acquire the current positioning data of the autonomous vehicle and the path planning result corresponding to the current positioning data of the autonomous vehicle, wherein the path planning result includes multiple alternative lane-level paths.

[0096] The second acquisition unit 220 is used to acquire the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle based on the current positioning data of the autonomous vehicle.

[0097] The first determining unit 230 is used to determine the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, wherein the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0098] The second determining unit 240 is used to determine the target lane-level path from multiple candidate lane-level paths based on the travel cost of each candidate lane-level path.

[0099] In some embodiments of this application, the second acquisition unit 220 is specifically used to: determine the predicted positioning data of the autonomous vehicle within a preset time period based on the current positioning data of the autonomous vehicle; acquire the corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predict the predicted positioning data of the obstacle within the preset time period based on the obstacle perception data.

[0100] In some embodiments of this application, the second acquisition unit 220 is specifically used to: establish a Frenet coordinate system corresponding to the alternative lane-level path based on the target lane in the alternative lane-level path; convert the current positioning data of the autonomous vehicle to the Frenet coordinate system corresponding to the alternative lane-level path; and determine the predicted positioning data of the autonomous vehicle within a preset time period based on the current positioning data of the autonomous vehicle in the Frenet coordinate system corresponding to the alternative lane-level path.

[0101] In some embodiments of this application, the second acquisition unit 220 is specifically used for: establishing a Frenet coordinate system corresponding to the alternative lane-level path based on the target lane in the alternative lane-level path; acquiring corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacle within a preset time period based on the obstacle perception data; and converting the predicted positioning data of the obstacle within the preset time period to the Frenet coordinate system corresponding to the alternative lane-level path to obtain the predicted positioning data of the obstacle corresponding to the alternative lane-level path.

[0102] In some embodiments of this application, the second acquisition unit 220 is specifically used to: determine the lane width of the target lane in the candidate lane-level path; and filter the predicted positioning data of obstacles in the Frenet coordinate system corresponding to the candidate lane-level path according to the lane width of the target lane in the candidate lane-level path to obtain the predicted positioning data of obstacles corresponding to the candidate lane-level path.

[0103] In some embodiments of this application, the first determining unit 230 is specifically used to: determine the longitudinal distance between the autonomous vehicle and the obstacle based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle; and determine the driving cost corresponding to each of the alternative lane-level paths based on the longitudinal distance between the autonomous vehicle and the obstacle.

[0104] In some embodiments of this application, the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle both include predicted positions at multiple prediction times. The first determining unit 230 is specifically used to: determine the longitudinal distance between the autonomous vehicle and the obstacle at each prediction time based on the predicted positions of the autonomous vehicle and the obstacle at each prediction time; and determine the driving cost of each alternative lane-level path using a preset weighted summation algorithm based on the longitudinal distance between the autonomous vehicle and the obstacle at each prediction time.

[0105] In some embodiments of this application, the current location data of the autonomous vehicle further includes the current speed and current acceleration. The first determining unit 230 is specifically used to: determine the current driving state of the autonomous vehicle based on the current speed and the current acceleration; and determine the driving cost of each alternative lane-level path based on the current driving state of the autonomous vehicle, the predicted location data of the autonomous vehicle, and the predicted location data of the obstacle.

[0106] It is understood that the path selection device for the autonomous vehicle described above can implement each step of the path selection method for the autonomous vehicle provided in the foregoing embodiments. The relevant explanations of the path selection method for the autonomous vehicle are applicable to the path selection device for the autonomous vehicle, and will not be repeated here.

[0107] Figure 3 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Please refer to it. Figure 3At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for other business operations.

[0108] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0109] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0110] The processor reads the corresponding computer program from non-volatile memory into main memory and then executes it, forming the path selection mechanism for the autonomous vehicle at the logical level. The processor executes the program stored in memory and specifically performs the following operations:

[0111] The current location data of the autonomous vehicle and the path planning results corresponding to the current location data of the autonomous vehicle are obtained, and the path planning results include multiple alternative lane-level paths.

[0112] Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle and the predicted location data of the obstacles are obtained.

[0113] Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the driving cost of each alternative lane-level path is determined, and the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0114] Based on the driving cost of each alternative lane-level path, the target lane-level path is determined from multiple alternative lane-level paths.

[0115] The above is as stated in this application. Figure 1 The method executed by the path selection device of the autonomous vehicle disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0116] The electronic device can also perform Figure 1 The method for executing the path selection device of an autonomous vehicle, and the implementation of the path selection device of an autonomous vehicle in... Figure 1 The functions of the embodiments shown are not described in detail here.

[0117] This application also proposes a computer-readable storage medium that stores one or more programs, the programs including instructions that, when executed by an electronic device including multiple applications, enable the electronic device to perform... Figure 1 The method executed by the path selection device of the autonomous vehicle in the illustrated embodiment is specifically used to perform:

[0118] The current location data of the autonomous vehicle and the path planning results corresponding to the current location data of the autonomous vehicle are obtained, and the path planning results include multiple alternative lane-level paths.

[0119] Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle and the predicted location data of the obstacles are obtained.

[0120] Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the driving cost of each alternative lane-level path is determined, and the driving cost is used to characterize the congestion level of each alternative lane-level path.

[0121] Based on the driving cost of each alternative lane-level path, the target lane-level path is determined from multiple alternative lane-level paths.

[0122] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0123] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0124] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0125] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0126] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0127] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0128] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0129] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0130] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of this application should be included within the scope of the claims of this application.

Claims

1. A path selection method for an autonomous vehicle, wherein, The method includes: The current location data of the autonomous vehicle and the path planning results corresponding to the current location data of the autonomous vehicle are obtained, and the path planning results include multiple alternative lane-level paths. Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle and the predicted location data of the obstacles are obtained. Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the driving cost of each alternative lane-level path is determined, and the driving cost is used to characterize the congestion level of each alternative lane-level path. Based on the driving cost of each alternative lane-level path, the target lane-level path is determined from multiple alternative lane-level paths. Both the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle include predicted positions at multiple prediction times. Determining the travel cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle includes: Based on the predicted positions of the autonomous vehicle and the obstacle at each predicted time, the longitudinal distance between the autonomous vehicle and the obstacle at each predicted time is determined. Based on the longitudinal distance between the autonomous vehicle and the obstacle at each time, calculate the driving cost of the alternative lane-level path at each time. Based on the driving cost of the candidate lane-level path at each time point and the corresponding weight at each time point, the driving cost of each candidate lane-level path is determined using a preset weighted summation algorithm. The weight of each moment gradually decreases over time.

2. The path selection method for autonomous vehicles as described in claim 1, wherein, The step of obtaining the predicted location data of the autonomous vehicle and the predicted location data of the obstacles based on the current location data of the autonomous vehicle includes: Based on the current location data of the autonomous vehicle, the predicted location data of the autonomous vehicle within a preset time period is determined. Based on the current location data of the autonomous vehicle, corresponding obstacle perception data is obtained, and the predicted location data of the obstacle within a preset time period is predicted based on the obstacle perception data.

3. The path selection method for autonomous vehicles as described in claim 2, wherein, The step of determining the predicted location data of the autonomous vehicle within a preset time period based on the current location data of the autonomous vehicle includes: Establish the Frenet coordinate system corresponding to the candidate lane-level path based on the target lane in the candidate lane-level path; The current location data of the autonomous vehicle is converted to the Frenet coordinate system corresponding to the alternative lane-level path; Based on the current positioning data of the autonomous vehicle in the Frenet coordinate system corresponding to the alternative lane-level path, the predicted positioning data of the autonomous vehicle within a preset time period is determined.

4. The path selection method for autonomous vehicles as described in claim 2, wherein, The step of acquiring corresponding obstacle perception data based on the current positioning data of the autonomous vehicle, and predicting the predicted positioning data of the obstacles within a preset time period based on the obstacle perception data includes: Establish the Frenet coordinate system corresponding to the candidate lane-level path based on the target lane in the candidate lane-level path; Based on the current location data of the autonomous vehicle, the corresponding obstacle perception data is obtained, and the predicted location data of the obstacle within a preset time period is predicted based on the obstacle perception data. The predicted location data of the obstacles within the preset time period are converted to the Frenet coordinate system corresponding to the alternative lane-level path to obtain the predicted location data of the obstacles corresponding to the alternative lane-level path.

5. The path selection method for autonomous vehicles as described in claim 4, wherein, The step of converting the predicted location data of the obstacles within the preset time period to the Frenet coordinate system corresponding to the candidate lane-level path, so as to obtain the predicted location data of the obstacles corresponding to the candidate lane-level path, includes: Determine the lane width of the target lane in the candidate lane-level paths; Based on the lane width of the target lane in the candidate lane-level path, the predicted location data of obstacles in the Frenet coordinate system corresponding to the candidate lane-level path are filtered to obtain the predicted location data of obstacles corresponding to the candidate lane-level path.

6. The path selection method for an autonomous vehicle as described in claim 1, wherein, The step of determining the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles includes: Based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, the longitudinal distance between the autonomous vehicle and the obstacle is determined; The driving cost corresponding to each of the alternative lane-level paths is determined based on the longitudinal distance between the autonomous vehicle and the obstacle.

7. A path selection device for an autonomous vehicle, wherein, The device includes: The first acquisition unit is used to acquire the current positioning data of the autonomous vehicle and the path planning result corresponding to the current positioning data of the autonomous vehicle, wherein the path planning result includes multiple alternative lane-level paths. The second acquisition unit is used to acquire the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacles based on the current positioning data of the autonomous vehicle. The first determining unit is used to determine the driving cost of each alternative lane-level path based on the predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle, wherein the driving cost is used to characterize the congestion level of each alternative lane-level path. The second determining unit is used to determine the target lane-level path from multiple candidate lane-level paths based on the travel cost of each candidate lane-level path. The predicted positioning data of the autonomous vehicle and the predicted positioning data of the obstacle both include predicted positions at multiple prediction times, and the first determining unit is specifically used for: Based on the predicted positions of the autonomous vehicle and the obstacle at each predicted time, the longitudinal distance between the autonomous vehicle and the obstacle at each predicted time is determined. Based on the longitudinal distance between the autonomous vehicle and the obstacle at each time, calculate the driving cost of the alternative lane-level path at each time. Based on the driving cost of the candidate lane-level path at each time point and the corresponding weight at each time point, the driving cost of each candidate lane-level path is determined using a preset weighted summation algorithm. The weight of each moment gradually decreases over time.

8. An electronic device, comprising: processor; as well as A memory configured to store computer-executable instructions, which, when executed, cause the processor to perform the path selection method for the autonomous vehicle according to any one of claims 1 to 6.

9. A computer-readable storage medium storing one or more programs, which, when executed by an electronic device including a plurality of applications, cause the electronic device to perform the path selection method for an autonomous vehicle according to any one of claims 1 to 6.

Images