Self-vehicle longitudinal decision-making method and system and electronic equipment

Abstract

The invention provides a self-vehicle longitudinal decision-making method, which belongs to the field of intelligent driving and comprises the following steps of: calculating an intersection point of tracks of a self-vehicle and a target vehicle and a time difference of arriving at the intersection point; the target vehicles are classified according to the time difference, the acceleration range of the vehicle under the corresponding condition is obtained, the longitudinal decision-making method for the vehicle can accurately judge the robbing and avoiding relation between the vehicle and the target vehicles, such as the overtaking of the target vehicle, the following of the target vehicle or the avoidance of collision, comprehensively considers the situation of multiple target vehicles, and improves the accuracy of the judgment result. The feasible acceleration range of the vehicle is reasonably determined, so that the driving safety and efficiency are improved, and the type of the target vehicle and the estimated acceleration range of the vehicle are given to downstream path planning reference for use.

Description

Technical Field

[0001] This invention relates to the field of intelligent driving technology, and in particular to a longitudinal decision-making method, system and electronic device for autonomous vehicles. Background Technology

[0002] In intelligent transportation and autonomous driving scenarios, accurately analyzing the relationships between vehicles and determining reasonable driving strategies for autonomous vehicles is crucial. Existing technologies often struggle to accurately determine the leading or following relationships between vehicles in complex multi-vehicle traffic scenarios, leading to unreasonable driving decisions by the autonomous vehicle. For example, in scenarios involving multiple vehicles merging or following each other, problems such as increased collision risk or low driving efficiency may occur.

[0003] In existing technologies, some vehicle driving strategy determination methods are based solely on the vehicle's current speed and distance information. For example, by acquiring the real-time distance and relative speed between the target vehicle and the vehicle itself using onboard radar, the vehicle decelerates when the distance is less than a certain threshold and the relative speed is greater than zero (indicating the target vehicle is approaching); conversely, it accelerates or maintains a constant speed when the distance is greater than a certain threshold and the relative speed is less than zero (indicating the target vehicle is moving away). This determination method is simple and direct, relying only on a single data dimension at the current moment.

[0004] This method ignores vehicle trajectory information and cannot accurately predict whether future trajectory conflicts will occur. For example, when the vehicle and the target vehicle are currently far apart but their trajectories have a potential intersection, current technology cannot make reasonable adjustments to the driving strategy in advance, which can easily lead to traffic accidents or traffic congestion. Moreover, judging solely based on current distance and speed does not take into account dynamic changes in the vehicle, such as acceleration and heading angle, resulting in inaccurate and incomplete judgments. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art, and proposes a longitudinal decision-making method and system for autonomous vehicles.

[0006] In a first aspect, the present invention provides a vehicle longitudinal decision-making method, comprising:

[0007] Calculate the intersection point of the trajectories of the vehicle and the target vehicle, as well as the time difference between reaching the intersection point;

[0008] The target vehicles are classified according to the time difference, and the acceleration range of the vehicles under the corresponding conditions is obtained.

[0009] In some embodiments, the steps of calculating the intersection point of the two vehicle trajectories and the time difference in reaching the intersection point include:

[0010] The predicted trajectory of the target vehicle is converted to the Frenet coordinate system of the vehicle trajectory. A coarse search is first performed to determine whether there are any intersections and the approximate location of the intersections.

[0011] When there are intersections, perform a fine search to calculate the coordinates of the intersections;

[0012] Calculate the time to reach the intersection point based on the position of the intersection point on the trajectories of the vehicle and the target vehicle, as well as the speed and acceleration of the two vehicles.

[0013] In some embodiments, the following steps are included.

[0014] Sampling points s are performed progressively along the predicted trajectory of the target vehicle, with a sampling step size of s. sample_dist Then the length s after the i-th sampling i =i·s sample_dist ;

[0015] For each sampling point si, a topological function T(si) is calculated. The corresponding topological symbol is determined by the topological function, and the topological symbols are concatenated to form a string T. When d(s)≥δ, the topological symbol is positive, and when d(s)≤-δ, the topological symbol is negative. At this time, the predicted trajectory of the target vehicle and the trajectory of the vehicle do not intersect at the sampling point s. When |d(s)|<δ, the predicted trajectory of the target vehicle and the trajectory of the vehicle intersect at the sampling point s. Here, d(s) is the lateral distance between the predicted trajectory of the target vehicle and the trajectory of the vehicle at the sampling point, and δ is the preset minimum intersection threshold.

[0016] And when s i Sampling stops when the length of string T exceeds S_max or reaches max_size_of_topology.

[0017] In some embodiments, during the coarse search process, when the topological symbol of two adjacent sampling points changes, a fine search is performed between these two points; and when the first sampling point s is found such that |d(s)|<δ, then sampling point s is the intersection point between the predicted trajectory of the target vehicle and the trajectory of the vehicle.

[0018] In some of these embodiments,

[0019] When the predicted trajectory of the target vehicle intersects with the trajectory of the vehicular vehicle, calculate the time t = s_M / when moving to the intersection point. s_M is the longitudinal distance to the intersection point in the Frenet coordinate system. This represents the current longitudinal velocity.

[0020] In some of these embodiments,

[0021] Also includes:

[0022] definition t=tego- t agent , where t ego Let t be the time it takes for the vehicle to move to the intersection. agent The time it takes for the target vehicle to move to the intersection point;

[0023] If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max The maximum time threshold is set.

[0024] If there is an intersection, and t leader > t>t max The target vehicle is classified as the leader; leader The target vehicle should be used as the threshold for the leader vehicle.

[0025] If there is an intersection, and -t max < t<-t follower The target vehicle is classified as a follower; follower The target vehicle should be used as the threshold for the self-vehicle followers;

[0026] If there is an intersection, and -t follower < t <t leader If the relationship between the target vehicle and the vehicle that is vying for right-of-way is unclear, it is temporarily designated as a risk target.

[0027] In some of these embodiments,

[0028] When the target vehicle is a risky target, a sliding window is used to update the data within the window range in real time, and the frame capacity n of the sliding window is set.

[0029] Calculate the integer score r for each frame within the sliding window. i And it is set that if the vehicle is in front of the target vehicle, then r i =1; if the vehicle is behind the target vehicle, r i =-1; if a collision risk is predicted, r i =-10;

[0030] The weights are distributed normally, where μ is the center of the current window, and σ is set empirically. The specific weight formula is as follows:

[0031] The standard deviation σ is set based on experience;

[0032] Calculate the weighted total score S: ;

[0033] Classification and acceleration constraints are based on the value of S:

[0034] If S>0: The target vehicle is the follower, calculate the collision avoidance deceleration a. min And the vehicle's acceleration a self >a min

[0035] If S < 0 and S > = -n: the target vehicle is the leader, calculate the collision avoidance acceleration a. max And the vehicle's acceleration a self max

[0036] If S <- n: The target vehicle is avoid_collision and the vehicle decelerates or accelerates to avoid collision.

[0037] In some of these embodiments,

[0038] The step of taking the intersection of all acceleration constraints to obtain the feasible acceleration range of the vehicle includes:

[0039] For each target, obtain the acceleration constraint interval according to its classification:

[0040] By finding the intersection of the constraints of all objectives, the feasible acceleration range of the vehicle can be obtained. ;

[0041] If the intersection is empty, the default range is used.

[0042] Secondly, the present invention also provides a vehicle longitudinal decision-making system, comprising:

[0043] The calculation unit is used to calculate the intersection point of the trajectories of the vehicular and target vehicles and the time difference between reaching the intersection point;

[0044] The control unit is used to classify the target vehicle according to the time difference and obtain the acceleration range of the vehicle under the corresponding conditions.

[0045] Thirdly, the present invention also provides an electronic device, comprising:

[0046] One or more processors;

[0047] Memory, used to store one or more programs;

[0048] When the one or more programs are executed by the one or more processors, the one or more processors implement the method.

[0049] Fourthly, the present invention also provides a computer-readable medium on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps in any of the methods described.​

[0050] The longitudinal decision-making method for autonomous vehicles provided by this invention calculates the intersection point of the trajectories of the autonomous vehicle and the target vehicle, as well as the time difference in reaching the intersection point. Based on the time difference, the target vehicle is classified, and the acceleration range of the autonomous vehicle under the corresponding conditions is obtained. The longitudinal decision-making method for autonomous vehicles provided by this invention can accurately determine the yielding relationship between the autonomous vehicle and the target vehicle, such as whether to overtake the target vehicle (leader), follow the target vehicle (follower), or avoid a collision. It also comprehensively considers the situation of multiple target vehicles and reasonably determines the feasible acceleration range of the autonomous vehicle to improve driving safety and efficiency. The type of target vehicle and the estimated acceleration range of the autonomous vehicle are provided to the downstream path planning reference. Attached Figure Description

[0051] Figure 1 This is a schematic diagram illustrating the steps of an embodiment of the vehicle longitudinal decision-making method of the present invention;

[0052] Figure 2 This is a schematic diagram illustrating the steps of an embodiment of the present invention in classifying target vehicles based on the time difference and obtaining the acceleration range of the vehicle under corresponding conditions;

[0053] Figure 3 This is a schematic diagram of an embodiment of a vehicle longitudinal decision-making system according to the present invention;

[0054] Figure 4 This is a schematic diagram of the structure of an embodiment of the electronic device of the present invention. Detailed Implementation

[0055] To enable those skilled in the art to better understand the technical solutions of the present invention, exemplary embodiments of the present invention are described below in conjunction with the accompanying drawings, including various details of the embodiments of the present invention to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0056] Where there is no conflict, the various embodiments of the present invention and the features thereof may be combined with each other.

[0057] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.

[0058] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Terms such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.

[0059] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having the meaning consistent with their meaning in the context of the relevant art and the invention, and will not be interpreted as having an idealized or overly formal meaning unless expressly so defined herein.

[0060] In the technical solution of this invention, the collection, storage, use, processing, transmission, provision, and disclosure of user personal information all comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example: appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely locating a specific individual.

[0061] In related technologies, particularly in intelligent transportation and autonomous driving scenarios, accurately analyzing the relationships between vehicles and determining reasonable driving strategies for autonomous vehicles is crucial. Existing technologies often struggle to accurately determine the leading or following relationships between vehicles in complex multi-vehicle traffic scenarios, leading to unreasonable driving decisions. For example, in scenarios involving multiple vehicles merging or following each other, this can result in increased collision risks or decreased driving efficiency.

[0062] In this invention, the input to the target vehicle is obtained from prediction and perception fusion, while the input to the driver vehicle is obtained from planning and chassis information. The output classifies each target vehicle with a label: leader indicates that the target vehicle will be in front of the driver vehicle in the future and needs to be followed; follower indicates that the target vehicle will be behind the driver vehicle in the future and needs to be overtaken; avoid_collision indicates that there is a high probability of collision in the future and needs to be slowed down.

[0063]

[0064] To address at least one of the technical problems existing in the aforementioned related technologies, the present invention provides a longitudinal decision-making method for autonomous vehicles. Figure 1 A flowchart of a vehicle longitudinal decision-making method provided in an embodiment of the present invention includes the following steps:

[0065] Step S10: Calculate the intersection point of the trajectories of the vehicle and the target vehicle, as well as the time difference between reaching the intersection point.

[0066] In this embodiment, the steps of calculating the intersection point of the two vehicle trajectories and the time difference in reaching the intersection point include:

[0067] The predicted trajectory of the target vehicle is converted to the Frenet coordinate system of the vehicle trajectory. A coarse search is first performed to determine whether there are any intersections and the approximate location of the intersections.

[0068] When there are intersections, perform a fine search to calculate the coordinates of the intersections;

[0069] Calculate the time to reach the intersection point based on the position of the intersection point on the trajectories of the vehicle and the target vehicle, as well as the speed and acceleration of the two vehicles.

[0070] Specifically, the following steps are included:

[0071] Sampling points s are performed progressively along the predicted trajectory of the target vehicle, with a sampling step size of s. sample_dist Then the length s after the i-th sampling i =i·s sample_dist ;

[0072] At each sampling point s i Calculate the topological function T(s) i The topological symbol is determined by the topological function, and the topological symbols are connected in sequence to form a string T. When d(s)≥δ, the topological symbol is positive, and when d(s)≤-δ, the topological symbol is negative. At this time, the predicted trajectory of the target vehicle and the trajectory of the vehicle do not intersect at the sampling point s. However, when |d(s)|<δ, the predicted trajectory of the target vehicle and the trajectory of the vehicle intersect at the sampling point s. Here, d(s) is the lateral distance between the predicted trajectory of the target vehicle and the trajectory of the vehicle at the sampling point, and δ is the preset minimum intersection threshold.

[0073] And when s i Sampling stops when the length of string T exceeds S_max or reaches max_size_of_topology.

[0074] Furthermore, during the coarse search process, when the topological sign of two adjacent sampling points changes, a fine search is performed between these two points;

[0075] Furthermore, when the first sampling point s is found such that |d(s)|<δ, then sampling point s is the intersection point between the predicted trajectory of the target vehicle and the trajectory of the vehicle.

[0076] It should be noted that since both the planned trajectory points of the autonomous vehicle and the predicted trajectory points of the target vehicle are discrete two-dimensional coordinates in a Cartesian coordinate system, and are not fitted using a function curve, the intersection point cannot be directly calculated. This method first transforms the predicted trajectory of the target vehicle into the Frenet coordinate system of the autonomous vehicle's trajectory, performs a coarse search to determine if there is an intersection point and its approximate location; then performs a fine search to calculate the coordinates of the intersection point; finally, based on the position of the intersection point on the trajectories of the autonomous vehicle and the target vehicle, and the velocities and accelerations of the two vehicles, the time to reach the intersection point is calculated for each. The core logic is given below, and its main process is described using mathematical formulas.

[0077] Suppose there are two reference curves: the current curve (i.e., the vehicle's planned trajectory) and the curve to be compared (Target vehicle predicted trajectory). The objective of the function is to follow... Sampling in the positive direction (along the s direction), calculating the lateral deviation d(s) between the two curves (i.e., At a certain point The coordinates (d-coordinates) in the coordinate system are used to classify the data according to the size of d(s), and a "topological string" T is constructed to map to a specific relation type.

[0078] Sampling and lateral bias definitions: Let Let s be the parameter at any point on the t. The Cartesian coordinates at point s are: =Frenet2Cartesian_ (s,0). The... Switch to The Frenet coordinate system is denoted as: F(s) = (x(s), d(s)) = Cartesian²Frenet_ ( ).

[0079] Here, d(s) represents the lateral distance between the two curves. The sampling step size is s. sample_dist =3.0m, and perform discrete sampling within s∈[0,S_max] (S_max is the longest distance of the trajectory).

[0080] Define a topological sign function: Define a threshold δ = merge_threshold (unit: m), where two points are considered to have intersecting trajectories if the distance between them is less than this threshold. For any sampling point s, we have

[0081]

[0082] Constructing the topology string:

[0083] Starting from s=0, with a step size s sample_dist Sampling, denoted as s i =i·s sample_dist i=0,1,2,…

[0084] At each sampling point s i Calculate T(s) i ), and concatenate them sequentially to form a string T (e.g., "o", "+o", "+o-", etc.). Sampling continues until one of two conditions is met:

[0085] ① The maximum topology length, max_size_of_topology, is reached; ②s i It exceeds S_max.

[0086] Additionally, if a change from non-'o' to 'o' occurs within a certain interval (i.e., the topological sign changes), the function will [s i -s sample_dist ,s i Within the range, a finer step size (0.2m) is used to search for the first sampling point s that satisfies |d(s)|<δ, and the corresponding value at this point is... As a mergepoint.

[0087] Final Relationship Determination: Based on the obtained topological string T, map it to a specific relation type. The mapping relationship is defined in the `relation_table`, for example:

[0088]

[0089] Merging time calculation:

[0090] Constant speed model: Assuming the vehicle travels at a constant speed V0, the time t to reach the intersection point is t = s_M / V0, where s_M is the longitudinal distance to the intersection point in the Frenet coordinate system.

[0091] It should be noted that: here we assume that the trajectory of the vehicle C1 and the trajectory of the target vehicle C2 are both continuous and smooth, and that the lateral distance can be accurately calculated through Frenet transformation.

[0092] The above scheme balances computational efficiency and accuracy through a hierarchical search strategy, and can effectively detect intersections and calculate relevant time parameters in the case of discrete trajectory points.

[0093] Step S20: Classify the target vehicles according to the time difference and obtain the acceleration range of the vehicles under the corresponding conditions.

[0094] Please see Figure 2 The step of classifying target vehicles based on the time difference and obtaining the acceleration range of the vehicle under the corresponding conditions includes:

[0095] Step S21: Coarse classification of targets: Based on the intersection point and Δt, classify the target vehicles into ignore, leader, follower, or risk targets.

[0096] definition t=t ego- t agent , where t ego Let t be the time it takes for the vehicle to move to the intersection. agent The time it takes for the target vehicle to move to the intersection point.

[0097] In this embodiment, the step of classifying target vehicles into ignore, leader, follower, or risk targets based on the intersection point and Δt includes:

[0098] If there is no intersection, classify the target vehicle as "ignore".

[0099] If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max This is the maximum time threshold that is set.

[0100] In both of these scenarios, the target is considered to have little impact on the vehicle's driving and is therefore ignored.

[0101] If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max The maximum time threshold is set.

[0102] If there is an intersection, and t leader > t>t max The target vehicle is classified as the leader; leader The target vehicle should be used as the threshold for the leader vehicle.

[0103] If there is an intersection, and -t max < t<-t follower The target vehicle is classified as a follower; follower The target vehicle should be used as the threshold for the self-vehicle followers;

[0104] If there is an intersection, and -t follower < t <t leaderIf the relationship between the target vehicle and the vehicle that is vying for right-of-way is unclear, it is temporarily designated as a risk target.

[0105] Step S22: For risk targets, perform detailed classification: use a sliding window to record the scores of historical frames, calculate the weighted total score S, further classify according to the value of S, and calculate acceleration constraints.

[0106] In this embodiment,

[0107] When the target vehicle is a risky target, a sliding window is used to update the data within the window range in real time. The frame capacity n of the sliding window is set; when n=30, it means that the sliding window records a maximum of 30 frames and stores a minimum of 10 frames. For each frame i (i=1,2,...,30), an integer score r is calculated based on the prediction. i ;

[0108] Calculate the integer score r for each frame within the sliding window. i And it is set that if the vehicle is in front of the target vehicle, then r i =1; if the vehicle is behind the target vehicle, r i =-1; if a collision risk is predicted, r i =-10;

[0109] The weights are distributed normally, where μ is the center of the current window, and σ is set empirically. The specific weight formula is as follows:

[0110] The standard deviation σ is set based on experience;

[0111] Calculate the weighted total score S: ;

[0112] Classification and acceleration constraints are based on the value of S:

[0113] If S>0: The target vehicle is the follower, calculate the collision avoidance deceleration a. min And the vehicle's acceleration a self >a min

[0114] If S < 0 and S > = -30: the target vehicle is the leader, calculate the anti-collision acceleration a. max And the vehicle's acceleration a self max

[0115] If S < -30: The target vehicle is avoid_collision and the vehicle decelerates or accelerates to avoid collision.

[0116] Specifically, the minimum acceleration *a* required to avoid a collision is calculated through acceleration. acc ;​

[0117] Calculate the minimum deceleration 'a' required to avoid a collision by slowing down. deacc ;

[0118] And the feasible acceleration of the vehicle must satisfy [a] self >a acc or a self deacc ]

[0119] Since there are usually multiple target vehicles on the road, the final output acceleration range is... It will take the intersection of the acceleration ranges of each target. If a reasonable acceleration range cannot be output, the default range will be output.

[0120] Since the time to reach the intersection point is known, different accelerations can be calculated by setting different safety distances. The details of how each acceleration is calculated will not be elaborated here.

[0121] This scheme improves decision-making accuracy while ensuring real-time performance through a coarse-fine two-level classification mechanism, making it particularly suitable for multi-objective interactive scenarios in complex and dynamic traffic environments.

[0122] The longitudinal decision-making method for autonomous vehicles provided by this invention calculates the intersection point of the trajectories of the autonomous vehicle and the target vehicle, as well as the time difference in reaching the intersection point. Based on the time difference, the target vehicle is classified, and the acceleration range of the autonomous vehicle under the corresponding conditions is obtained. The longitudinal decision-making method for autonomous vehicles provided by this invention can accurately determine the yielding relationship between the autonomous vehicle and the target vehicle, such as whether to overtake the target vehicle (leader), follow the target vehicle (follower), or avoid a collision. It also comprehensively considers the situation of multiple target vehicles and reasonably determines the feasible acceleration range of the autonomous vehicle to improve driving safety and efficiency. The type of target vehicle and the estimated acceleration range of the autonomous vehicle are provided to the downstream path planning reference.

[0123] Please see Figure 3 The present invention also provides a vehicle longitudinal decision-making system, which is applied to the vehicle longitudinal decision-making method provided in the above embodiments, specifically including: a computing unit and a control unit.

[0124] The calculation unit is used to calculate the intersection point of the trajectories of the vehicular and target vehicles and the time difference between reaching the intersection point.

[0125] The predicted trajectory of the target vehicle is converted to the Frenet coordinate system of the vehicle trajectory. A coarse search is first performed to determine whether there are any intersections and the approximate location of the intersections.

[0126] When there are intersections, perform a fine search to calculate the coordinates of the intersections;

[0127] ​Calculate the time to reach the intersection point based on the position of the intersection point on the trajectories of the vehicle and the target vehicle, as well as the speed and acceleration of the two vehicles.

[0128] Specifically, the following steps are included:

[0129] Sampling points s are performed progressively along the predicted trajectory of the target vehicle, with a sampling step size of s. sample_dist Then the length s after the i-th sampling i =i·s sample_dist ;

[0130] For each sampling point si, a topological function T(si) is calculated. The corresponding topological symbol is determined by the topological function, and the topological symbols are concatenated to form a string T. When d(s)≥δ, the topological symbol is positive, and when d(s)≤-δ, the topological symbol is negative. At this time, the predicted trajectory of the target vehicle and the trajectory of the vehicle do not intersect at the sampling point s. When |d(s)|<δ, the predicted trajectory of the target vehicle and the trajectory of the vehicle intersect at the sampling point s. Here, d(s) is the lateral distance between the predicted trajectory of the target vehicle and the trajectory of the vehicle at the sampling point, and δ is the preset minimum intersection threshold.

[0131] And when s i Sampling stops when the length of string T exceeds S_max or reaches max_size_of_topology.

[0132] Furthermore, during the coarse search process, when the topological symbol of two adjacent sampling points changes, a fine search is performed between these two points; and when the first sampling point s is found such that |d(s)|<δ, then sampling point s is the intersection point between the predicted trajectory of the target vehicle and the trajectory of the vehicle.

[0133] The control unit is used to classify the target vehicle according to the time difference and obtain the acceleration range of the vehicle under the corresponding conditions.

[0134] When the predicted trajectory of the target vehicle intersects with the trajectory of the vehicular vehicle, the time t = s_M / V0 is calculated to move to the intersection point, where s_M is the longitudinal distance to the intersection point in the Frenet coordinate system and V0 is the current longitudinal speed.

[0135] And definition t=t ego- t agent , where t ego Let t be the time it takes for the vehicle to move to the intersection. agent The time it takes for the target vehicle to move to the intersection point.

[0136] In this embodiment, the step of classifying target vehicles into ignore, leader, follower, or risk targets based on the intersection point and Δt includes:

[0137] If there is no intersection, classify the target vehicle as "ignore".

[0138] If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max This is the maximum time threshold that is set.

[0139] In both of these scenarios, the target is considered to have little impact on the vehicle's driving and is therefore ignored.

[0140] If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max The maximum time threshold is set.

[0141] If there is an intersection, and t leader > t>t max The target vehicle is classified as the leader; leader The target vehicle should be used as the threshold for the leader vehicle.

[0142] If there is an intersection, and -t max < t<-t follower The target vehicle is classified as a follower; follower The target vehicle should be used as the threshold for the self-vehicle followers;

[0143] If there is an intersection, and -t follower < t <t leader If the relationship between the target vehicle and the vehicle that is vying for right-of-way is unclear, it is temporarily designated as a risk target.

[0144] For risk targets, further classification is performed: a sliding window is used to record the scores of historical frames, a weighted total score S is calculated, further classification is performed based on the value of S, and acceleration constraints are calculated.

[0145] In this embodiment,

[0146] When the target vehicle is a risky target, a sliding window is used to update the data within the window range in real time, and the frame capacity n of the sliding window is set.

[0147] Calculate the integer score r for each frame within the sliding window. i And it is set that if the vehicle is in front of the target vehicle, then r i =1; if the vehicle is behind the target vehicle, r i =-1; if a collision risk is predicted, r i =-10;

[0148] The weights are distributed normally, where μ is the center of the current window, and σ is set empirically. The specific weight formula is as follows:

[0149] The standard deviation σ is set based on experience;

[0150] Calculate the weighted total score S: ;

[0151] Classification and acceleration constraints are based on the value of S:

[0152] If S>0: The target vehicle is the follower, calculate the collision avoidance deceleration a. min And the vehicle's acceleration a self >a min

[0153] If S < 0 and S > = -n: the target vehicle is the leader, calculate the collision avoidance acceleration a. max And the vehicle's acceleration a self max

[0154] If S <- n: The target vehicle is avoid_collision and the vehicle decelerates or accelerates to avoid collision.

[0155] Specifically, the minimum acceleration (a) required to avoid a collision is calculated by accelerating. acc );

[0156] Calculate the minimum deceleration (a) required to avoid a collision by slowing down. deacc );

[0157] The feasible acceleration of the vehicle must satisfy [a] self >a acc or a self deacc ]

[0158] Since there are usually multiple target vehicles on the road, the final output acceleration range is... It will take the intersection of the acceleration ranges of each target. If a reasonable acceleration range cannot be output, the default range will be output.

[0159] Since the time to reach the intersection point is known, different accelerations can be calculated by setting different safety distances. The details of how each acceleration is calculated will not be elaborated here.

[0160] This scheme improves decision-making accuracy while ensuring real-time performance through a coarse-fine two-level classification mechanism, making it particularly suitable for multi-objective interactive scenarios in complex and dynamic traffic environments.

[0161] ​​The autonomous vehicle longitudinal decision-making system provided by this invention calculates the intersection point of the trajectory of the autonomous vehicle and the target vehicle, as well as the time difference in reaching the intersection point; classifies the target vehicle based on the time difference, and obtains the acceleration range of the autonomous vehicle under the corresponding conditions. The autonomous vehicle longitudinal decision-making method provided by this invention can accurately determine the right-of-way relationship between the autonomous vehicle and the target vehicle, such as whether to overtake the target vehicle (leader), follow the target vehicle (follower), or avoid a collision. It also comprehensively considers the situation of multiple target vehicles and reasonably determines the feasible acceleration range of the autonomous vehicle to improve driving safety and efficiency. The type of target vehicle and the estimated acceleration range of the autonomous vehicle are provided to downstream path planning for reference.

[0162] Based on the same inventive concept, embodiments of the present invention also provide an electronic device. Figure 4 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Figure 4 As shown, an embodiment of the present invention provides an electronic device including: one or more processors 101, a memory 102, and one or more I / O interfaces 103. The memory 102 stores one or more programs, which, when executed by the one or more processors, enable the one or more processors to implement any of the vehicle longitudinal decision-making methods described in the above embodiments; the one or more I / O interfaces 103 are connected between the processors and the memory, configured to enable information interaction between the processors and the memory.

[0163] The processor 101 is a device with data processing capabilities, including but not limited to a central processing unit (CPU); the memory 102 is a device with data storage capabilities, including but not limited to random access memory (RAM, more specifically SDRAM, DDR, etc.), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), and flash memory (FLASH); the I / O interface (read / write interface) 103 is connected between the processor 101 and the memory 102, and can realize information interaction between the processor 101 and the memory 102, including but not limited to a data bus (Bus).

[0164] In some embodiments, the processor 101, memory 102, and I / O interface 103 are interconnected via bus 104, and thus connected to other components of the computing device.

[0165] In some embodiments, the one or more processors 101 include a field-programmable gate array.

[0166] This invention also provides a computer-readable medium. The computer-readable medium stores a computer program, which, when executed by a processor, implements the steps of any of the vehicle longitudinal decision-making methods described in the above embodiments. The computer-readable storage medium may be volatile or non-volatile.

[0167] This invention also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code. When the computer-readable code is run in the processor of an electronic device, the processor in the electronic device executes the above-described vehicle longitudinal decision-making method.

[0168] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer storage media (or non-transitory media) and communication media (or transient media).

[0169] As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable program instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), flash memory or other memory technologies, portable compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable program instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0170] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0171] The computer program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the computer-readable program instructions. This electronic circuitry can execute the computer-readable program instructions to implement various aspects of the invention.

[0172] The computer program product described herein can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0173] Various aspects of the present invention are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should 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-readable program instructions.

[0174] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.

[0175] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.

[0176] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction, which contains one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those shown in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0177] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in conjunction with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in conjunction with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of the invention as set forth in the appended claims.

Claims

1. A longitudinal decision-making method for autonomous vehicles, characterized in that, It includes: Calculate the intersection point of the trajectories of the vehicle and the target vehicle, as well as the time difference between reaching the intersection point; The target vehicles are classified according to the time difference, and the acceleration range of the vehicles under the corresponding conditions is obtained.

2. The longitudinal decision-making method for a vehicle according to claim 1, characterized in that, The steps for calculating the intersection point of the two vehicle trajectories and the time difference between reaching the intersection point include: The predicted trajectory of the target vehicle is converted to the Frenet coordinate system of the vehicle trajectory. A coarse search is first performed to determine whether there are any intersections and the approximate location of the intersections. When there are intersections, perform a fine search to calculate the coordinates of the intersections; Calculate the time to reach the intersection point based on the position of the intersection point on the trajectories of the vehicle and the target vehicle, as well as the speed and acceleration of the two vehicles.

3. The longitudinal decision-making method for a vehicle according to claim 2, characterized in that, Includes the following steps Sampling points s are performed progressively along the predicted trajectory of the target vehicle, with a sampling step size of s. sample_dist Then the length s after the i-th sampling i =i·s sample_dist ; For each sampling point si, a topological function T(si) is calculated. The corresponding topological symbol is determined by the topological function, and the topological symbols are concatenated to form a string T. When d(s)≥δ, the topological symbol is positive, and when d(s)≤-δ, the topological symbol is negative. At this time, the predicted trajectory of the target vehicle and the trajectory of the vehicle do not intersect at sampling point s. When |d(s)|<δ, the predicted trajectory of the target vehicle and the trajectory of the vehicle intersect at sampling point s. Here, d(s) is the lateral distance between the predicted trajectory of the target vehicle and the trajectory of the vehicle at the sampling point, and δ is the preset minimum intersection threshold. And when s i Sampling stops when the length of string T exceeds S_max or reaches max_size_of_topology.

4. The vehicle longitudinal decision-making method according to claim 3, characterized in that, During the coarse search, when the topological sign of two adjacent sampling points changes, a fine search is performed between these two points. Furthermore, when the first sampling point s is found such that |d(s)|<δ, then sampling point s is the intersection point between the predicted trajectory of the target vehicle and the trajectory of the vehicle.

5. The longitudinal decision-making method for a vehicle according to claim 2, characterized in that, Also includes: When the predicted trajectory of the target vehicle intersects with the trajectory of the vehicular vehicle, the time t = s_M / V0 is calculated to move to the intersection point, where s_M is the longitudinal distance to the intersection point in the Frenet coordinate system and V0 is the current longitudinal speed.

6. The longitudinal decision-making method for a vehicle according to claim 5, characterized in that, Also includes: definition t=t ego -t agent , where t ego Let t be the time it takes for the vehicle to move to the intersection. agent The time it takes for the target vehicle to move to the intersection point; If there are intersections, and | t|>t max The target vehicle is classified as "ignore"; max The maximum time threshold is set. If there is an intersection, and t leader > t>t max The target vehicle is classified as the leader; leader The target vehicle should be used as the threshold for the leader vehicle. If there is an intersection, and -t max < t<-t follower The target vehicle is classified as a follower; follower The target vehicle should be used as the threshold for the self-vehicle follower; If there is an intersection, and -t follower < t <t leader If the relationship between the target vehicle and the vehicle that is vying for right-of-way is unclear, it is temporarily designated as a risk target.

7. The vehicle longitudinal decision-making method according to claim 6, characterized in that, When the target vehicle is a risky target, a sliding window is used to update the data within the window range in real time, and the frame capacity n of the sliding window is set. Calculate the integer score r for each frame within the sliding window. i And it is set that if the vehicle is in front of the target vehicle, then r i =1; if the vehicle is behind the target vehicle, r i =-1; if a collision risk is predicted, r i =-10; The weights are distributed normally, where μ is the center of the current window, and σ is set empirically. The specific weight formula is as follows: The standard deviation σ is set based on experience; Calculate the weighted total score S: ; Classification and acceleration constraints are based on the value of S: If S>0: The target vehicle is the follower, calculate the collision avoidance deceleration a. min And the vehicle's acceleration a self >a min If S < 0 and S > = -n: the target vehicle is the leader, calculate the collision avoidance acceleration a. max And the vehicle's acceleration a self max ​ If S <- n: The target vehicle is avoid_collision and the vehicle decelerates or accelerates to avoid collision.

8. The longitudinal decision-making method for a vehicle according to claim 7, characterized in that, The step of taking the intersection of all acceleration constraints to obtain the feasible acceleration range of the vehicle includes: For each target, obtain the acceleration constraint interval according to its classification: By finding the intersection of the constraints of all objectives, the feasible acceleration range of the vehicle can be obtained. ; If the intersection is empty, the default range is used.

9. A longitudinal decision-making system for autonomous vehicles, characterized in that, include: The calculation unit is used to calculate the intersection point of the trajectories of the vehicular and target vehicles and the time difference between reaching the intersection point; The control unit is used to classify the target vehicle according to the time difference and obtain the acceleration range of the vehicle under the corresponding conditions.

10. An electronic device, characterized in that, include: One or more processors; Memory, used to store one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 8.

Images