Passage planning system and method, electronic equipment and storage medium

Abstract

The invention discloses a traffic planning system and method, electronic equipment and a storage medium. The traffic planning method comprises the steps of obtaining type characteristics of interaction obstacles around a current vehicle; speculating the motion intention of the interaction obstacle according to the type characteristic of the interaction obstacle; and obtaining a passing decision of the current vehicle according to the motion intention. The motion intention of the dynamic obstacle is speculated according to the type characteristic of the dynamic obstacle, so that decision information is provided for the current vehicle, the problems of low passing efficiency, low safety and poor experience of the automatic driving vehicle at the intersection caused by the problems of frequent snub braking, rapid stopping and the like of the automatic driving vehicle are solved, the flexibility is high, the passing efficiency is high, the safety is high, and the practicability is high. And the experience feeling is improved.

Description

Technical Field

[0001] The present application relates to the technical field of vehicles, and specifically relates to a traffic planning system and its method, an electronic device, and a storage medium. Background Art

[0002] With the continuous development of autonomous driving technology, autonomous driving technology has gradually been applied to more complex urban scenarios. To implement autonomous driving technology in urban scenarios, in addition to considering the safety and comfort of ADC (Autonomous Driving Car) when running on the road, more attention should be paid to safety, traffic efficiency, and experience when passing through intersections. In related technologies, the traffic decision-making of ADC is single, with low flexibility, low traffic efficiency, low safety, and poor experience. Summary of the Invention

[0003] In view of this, the present application provides a traffic planning system and its method, an electronic device, and a storage medium to solve the problems in the prior art that the traffic decision-making of ADC is single, with low flexibility, low traffic efficiency, low safety, and poor experience.

[0004] To solve the above technical problems, the first technical solution provided by the present application is: to provide a traffic planning method, including:

[0005] Obtain the type characteristics of the interactive obstacles around the current vehicle;

[0006] Speculate the movement intention of the interactive obstacle according to the type characteristics of the interactive obstacle;

[0007] Obtain the traffic decision of the current vehicle according to the movement intention.

[0008] In one embodiment, before speculating the movement intention of the interactive obstacle according to the type characteristics of the interactive obstacle, it further includes:

[0009] Obtain the environmental information of the current vehicle;

[0010] The speculating the movement intention of the interactive obstacle according to the type characteristics of the interactive obstacle includes:

[0011] Speculate the movement intention of the interactive obstacle according to the type characteristics of the interactive obstacle itself, the traffic rules corresponding to the type characteristics, and the environmental information.

[0012] In one embodiment, before obtaining the traffic decision of the current vehicle according to the movement intention, it further includes:

[0013] Obtain the current running trajectory and acceleration of the interactive obstacle;

[0014] Obtain the predicted running trajectory of the interactive obstacle according to the current running trajectory, the acceleration, and the motion intention of the interactive obstacle.

[0015] In one embodiment, the obtaining the passing decision of the current vehicle according to the motion intention includes:

[0016] According to the motion state information and path planning information of the current vehicle at the current moment, and the predicted running trajectory of the interactive obstacle, select the interactive obstacle having an interactive relationship with the current vehicle from multiple interactive obstacles as the game target;

[0017] Set an interaction label for the game target; wherein the interaction label includes the running direction, road right information, and field of view information of the game target;

[0018] According to the interaction label, obtain the interaction method with the highest running probability between the current vehicle and the game target to obtain the passing decision of the current vehicle.

[0019] In one embodiment, the setting the interaction label for the game target includes:

[0020] Perform acceleration sampling on the current vehicle and the game target respectively;

[0021] Generate a first sampling space of the current vehicle and a second sampling space of the game target respectively according to the acceleration sampling;

[0022] Deduce the sampling trajectory according to the first sampling space to obtain the first longitudinal trajectory cluster of the current vehicle; deduce the sampling trajectory according to the second sampling space to obtain the second longitudinal trajectory cluster of the game target;

[0023] Perform one-to-one interaction on the first longitudinal trajectory cluster and the second longitudinal trajectory cluster to obtain the interaction cost of each interaction trajectory pair;

[0024] Select the interaction trajectory pair with the lowest interaction cost from the interaction costs of the interaction trajectory pairs as the interaction method with the highest running probability between the current vehicle and the game target.

[0025] In one embodiment, the interaction cost includes safety cost, comfort cost, efficiency cost, and road right cost; the efficiency cost is measured by the time difference between the current vehicle and the game target passing through the path coincidence point; wherein, the path coincidence point is the same path in the predicted running trajectories of the current vehicle and the game target;

[0026] Performing one-to-one interaction on the first longitudinal trajectory cluster and the second longitudinal trajectory cluster to obtain the interaction cost of each interaction trajectory pair includes:

[0027] Obtaining the safety cost of each interaction trajectory pair;

[0028] Obtaining the comfort cost of each interaction trajectory pair;

[0029] Obtaining the efficiency cost of each interaction trajectory pair;

[0030] Obtaining the right-of-way cost of each interaction trajectory pair;

[0031] By summing up the safety cost, the comfort cost, the efficiency cost and the right-of-way cost, obtaining the total cost of the interaction between the current vehicle and the game target.

[0032] In one embodiment, obtaining the efficiency cost of each interaction trajectory pair includes:

[0033] Obtaining the times when the first longitudinal trajectory cluster and the second longitudinal trajectory cluster respectively pass through the path coincidence point;

[0034] Obtaining the times when the current vehicle and the game target respectively pass through the path coincidence point with uniform acceleration at the current acceleration;

[0035] Obtaining a first difference, and a second difference between the maximum value and the minimum value of the first difference;

[0036] Based on the first difference and the second difference, and in combination with the penalty weight of the efficiency cost, obtaining the respective efficiency costs of the current vehicle and the game target.

[0037] In one embodiment, obtaining the interaction mode with the highest running probability between the current vehicle and the game target according to the interaction label to obtain the passing decision of the current vehicle includes:

[0038] In response to the current vehicle arriving at the path coincidence point earlier than the game target, the passing decision of the current vehicle is an overtaking game strategy;

[0039] In response to the current vehicle arriving at the path coincidence point later than the game target, the passing decision of the current vehicle is a yielding game strategy.

[0040] In one embodiment, after obtaining the passing decision of the current vehicle according to the motion intention, it includes:

[0041] Generate the driving trajectory information of the current vehicle according to the interaction method with the highest running probability between the current vehicle and the game target, and according to the traffic decision.

[0042] In one embodiment, the traffic planning method further includes:

[0043] Obtain the driving control information of the current vehicle according to the driving trajectory information of the current vehicle;

[0044] Send the driving control information to the execution end of the current vehicle; wherein the driving control information includes the driving acceleration and direction control information of the current vehicle.

[0045] To solve the above technical problems, the second technical solution provided by this application is: Provide a traffic planning system, including:

[0046] A perception and prediction module, configured to obtain the type characteristics of the interactive obstacles around the current vehicle;

[0047] A judgment module, configured to infer the motion intention of the interactive obstacle according to the type characteristics of the interactive obstacle;

[0048] A decision-making module, configured to obtain the traffic decision of the current vehicle according to the motion intention;

[0049] A control module, connected to the perception and prediction module, the judgment module and the decision-making module, configured to obtain the information of the corresponding module and send control instructions to the corresponding module.

[0050] To solve the above technical problems, the third technical solution provided by this application is: Provide an electronic device, including: a processor and a memory, the memory is connected to the processor and is used to store a computer program that can run on the processor; wherein, when the processor executes the computer program, the method described in any one of the above is implemented.

[0051] To solve the above technical problems, the fourth technical solution provided by this application is: Provide a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in any one of the above is implemented.

[0052] Advantages of the present application: Different from the prior art, the traffic planning method of the present application includes: obtaining the type characteristics of interactive obstacles around the current vehicle; inferring the motion intention of the interactive obstacles based on the type characteristics of the interactive obstacles; and obtaining the traffic decision of the current vehicle according to the motion intention. The present application infers the motion intention of dynamic obstacles based on their type characteristics, thereby providing decision-making information for the current vehicle, solving the problems of low traffic efficiency, low safety, and poor experience of autonomous vehicles at intersections caused by frequent point braking, awkward stopping, etc. of autonomous vehicles, with strong flexibility, high traffic efficiency, high safety, and improved experience at the same time. Description of the Drawings

[0053] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0054] Figure 1 is the overall flowchart of the traffic planning method provided by an embodiment of the present application;

[0055] Figure 2 is the flowchart of the steps before step S3 provided by an embodiment of the present application;

[0056] Figure 3 is Figure 1 or Figure 2 provides the flowchart of the sub-steps of step S3;

[0057] Figure 4 is Figure 3 provides the flowchart of the sub-steps of step S32;

[0058] Figure 5 is the schematic diagram of the predicted interaction trajectory between the current vehicle and the interactive obstacle provided by an embodiment of the present application;

[0059] Figure 6 is the schematic diagram of the field of view information judgment between the current vehicle and the interactive obstacle provided by an embodiment of the present application;

[0060] Figure 7 is the schematic table of the longitudinal acceleration sampling space of the current vehicle provided by an embodiment of the present application;

[0061] Figure 8 is the longitudinal acceleration change curve diagram of the current vehicle provided by an embodiment of the present application;

[0062] Figure 9It is a schematic diagram of a longitudinal trajectory cluster formed by the longitudinal deduction trajectory of the current vehicle or the interactive obstacle provided by an embodiment of the present application;

[0063] Figure 10 It is Figure 4 a flowchart of the sub-steps of step S324 provided;

[0064] Figure 11 It is Figure 10 a flowchart of the sub-steps of step S3243 provided;

[0065] Figure 12 It is Figure 3 a flowchart of the sub-steps of step S33 provided;

[0066] Figure 13 It is the overall flowchart of the traffic planning method provided by another embodiment of the present application;

[0067] Figure 14 It is a schematic diagram of the module connection of the traffic planning system provided by an embodiment of the present application;

[0068] Figure 15 It is a schematic diagram of the module connection of the traffic planning system provided by a further embodiment of the present application;

[0069] Figure 16 It is a schematic diagram of the connection and working process of the interactive game decision-making module provided by an embodiment of the present application;

[0070] Figure 17 It is a schematic diagram of the structure of the electronic device provided by an embodiment of the present application;

[0071] Figure 18 It is a schematic block diagram of the structure of the computer-readable storage medium provided by an embodiment of the present application.

[0072] Explanation of reference numerals:

[0073] 10. Current vehicle; 20. Interactive obstacle; 21. Game target; 30. Road; 40. Longitudinal trajectory cluster; 41. First longitudinal trajectory cluster; 42. Second longitudinal trajectory cluster; 100. Traffic planning system; 110. Perception and prediction module; 120. Judgment module; 130. Decision-making module; 140. Control module; 150. Planning module; 160. Game target screening module; 170. Vehicle actuator module; 200. Electronic device; 210. Processor; 220. Memory; 230. Peripheral device interface; 240. Radio frequency circuit; 250. Display screen; 260. Audio circuit; 270. Power supply; 300. Computer-readable storage medium. Detailed implementation manners

[0074] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0075] The terms "first" and "second" in the present application are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one of such features. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present application are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0076] Referring to "embodiments" herein means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those of ordinary skill in the art understand explicitly and implicitly that the embodiments described herein can be combined with other embodiments.

[0077] If the technical solution of this application involves personal information, before the product applying the technical solution of this application processes personal information, it has clearly informed the personal information processing rules and obtained the individual's independent consent. If the technical solution of this application involves sensitive personal information, before the product applying the technical solution of this application processes sensitive personal information, it has obtained the individual's separate consent and at the same time meets the requirements of "express consent". For example, at personal information collection devices such as cameras, clear and prominent signs are set to inform that the personal information collection range has been entered and personal information will be collected. If an individual voluntarily enters the collection range, it is regarded as consent to the collection of their personal information; or on the device for personal information processing, when the personal information processing rules are informed by obvious signs / information, personal authorization is obtained through pop-up messages or by asking the individual to upload their personal information themselves; among them, the personal information processing rules may include information such as the personal information processor, the purpose of personal information processing, the processing method, and the types of personal information processed.

[0078] During the research process of this application, it is found that when an autonomous driving car (ADC) passes through an intersection, by comprehensively considering the current information of the ADC (including path planning information, current speed, current acceleration), the information of obstacles (including predicted trajectory, current speed, current acceleration), and the interaction state between the ADC path and the obstacle trajectory, it is determined whether the ADC can pass through the current intersection safely and efficiently; when the ADC passes through an intersection, by comprehensively considering the current vehicle, the right-of-way information of obstacles, etc., the game strategy for the ADC to pass through the intersection is further determined to reduce the probability of danger and traffic jams caused when the ADC passes through the intersection. However, the traditional decision-making and planning based on a finite state machine are usually affected by the prediction accuracy of the target trajectory. When the motion intention of an obstacle is not clear, resulting in the predicted trajectory of the obstacle constantly jumping, there will be a situation where the predicted trajectory of the obstacle frequently invades the driving road of the autonomous driving vehicle. These situations will all lead to problems such as frequent point braking and awkward stops of the autonomous driving vehicle, resulting in low traffic efficiency, low safety, and poor experience of the autonomous driving vehicle at the intersection.

[0079] In response to the above problems existing when an autonomous driving vehicle passes through an intersection, there are some solutions in the prior art. However, most of them rely too much on the information of the predicted trajectory and combine a finite state machine to make decision-making and planning for the autonomous driving vehicle. This method has high requirements for upstream perception and prediction information and is feasible in an ideal simulation environment. If in a real intersection, perception and prediction cannot truly give the motion intention and trajectory information of obstacles, and the trajectory information has jumps. In response to this situation, the solution in the prior art is to process it through a finite state machine. The finite state machine can handle interaction problems in simple scenarios, but in the face of the complex intersection environment in China, the finite state machine is insufficient in completeness and often shows the phenomenon of being unable to attend to both sides.

[0080] Due to the instability of obstacle perception and trajectory prediction intentions in the real world, the method of relying too much on predicted trajectories for ADC games mentioned above will lead to instability of game results or even incorrect game results in the complex traffic flow at actual intersections; moreover, the method of intersection passing decision-making based on a finite state machine mentioned above is too rigid and lacks completeness, and cannot cope with complex intersections and traffic flows in China, posing potential safety hazards.

[0081] To solve the above problems, the present application provides a passing planning system, its method, an electronic device, and a storage medium.

[0082] Please refer to Figure 1 , the passing planning system 100 and its method provided by the present application are based on the current motion states of the interactive obstacle 20 and the current vehicle 10, sample possible motion trajectories according to different accelerations, then conduct a game by combining the environmental information where both interactive parties are located, screen out the optimal interactive method to generate the optimal passing decision for the current vehicle 10 for use in downstream path planning, generate the optimal passing decision through a cost function by fully considering the possible motion states of the obstacle and the current vehicle 10 and the environmental information, reduce situations such as point braking and awkward stops when the current vehicle 10 passes through the intersection on the premise of ensuring the safety of the current vehicle 10 and complying with traffic rules, improve the comfort and passing efficiency of the current vehicle 10 when passing through the intersection, and at the same time reduce the congestion at the intersection.

[0083] Please refer to Figure 1 , a passing planning method provided by an embodiment of the present application may include the following steps:

[0084] S1: Obtain the type characteristics of the interactive obstacle 20 around the current vehicle 10.

[0085] Specifically, the type characteristics of the interactive obstacle 20 may include motor vehicles, non-motor vehicles, and walking target objects, etc. Since the type characteristics of each interactive obstacle 20 correspond to different traffic rules, its possible motion intention can be inferred based on this type characteristic. It can be understood that the current vehicle 10 may be an autonomous driving vehicle (ADC).

[0086] S2: Infer the motion intention of the interactive obstacle 20 according to the type characteristics of the interactive obstacle 20.

[0087] Specifically, according to the type characteristics of each interaction obstacle 20, combined with the current position of the type characteristics and the corresponding traffic rules, the possible movement intention of the interaction obstacle 20 can be inferred. For example, at an intersection, if the type characteristic of the interaction obstacle 20 is a non-motor vehicle and the current position of the interaction obstacle 20 is on one side of the zebra crossing, then the movement intention of the interaction obstacle 20 may be to go straight through the zebra crossing; or, if the interaction obstacle 20 is in the left-turn waiting zebra crossing area, then the movement intention of the interaction obstacle 20 may be to turn left. Inferring the movement intention of the interaction obstacle 20 based on the type characteristics of the interaction obstacle 20 can improve the accuracy of inferring the movement intention of the interaction obstacle 20. The interaction method of inferring the movement intention of the interaction obstacle 20 has a higher prediction accuracy and is more flexible compared to the trajectory prediction of the interaction obstacle 20 only based on the running trajectory of the interaction obstacle 20 in the related art. The specific reasoning method can be through a game strategy, or through model prediction or other means for reasoning.

[0088] In one embodiment, before step S2 of inferring the movement intention of the interaction obstacle 20 according to the type characteristics of the interaction obstacle 20, the following may also be included:

[0089] Obtain the environmental information of the current vehicle 10.

[0090] Specifically, the environmental information may include the environment and traffic environment of the road 30 where the current vehicle 10 is located, such as whether it is a straight intersection, a crossroads or a T-junction, etc. The traffic environment may be whether straight, left-turn, right-turn or U-turn and other passing methods are available at the current intersection. For example, if the environment of the road 30 is a T-junction, going straight is not possible; if the current environment of the road 30 is a straight intersection and the corresponding traffic environment is a U-turn prohibition, then the movement intention of the interaction obstacle 20 (specifically a motor vehicle) to make a U-turn can be excluded. By obtaining the environmental information of the current vehicle 10, the prediction accuracy of the movement intention of the interaction obstacle 20 can be further improved.

[0091] Correspondingly, in this embodiment, step S2 of inferring the movement intention of the interaction obstacle 20 according to the type characteristics of the interaction obstacle 20 may include:

[0092] Infer the movement intention of the interaction obstacle 20 based on the type characteristics of the interaction obstacle 20 itself, the traffic rules corresponding to the type characteristics, and the environmental information.

[0093] Specifically, after obtaining the environmental information of the current vehicle 10, the type characteristics of the interactive obstacle 20 itself, and the corresponding traffic rules, the above information can be combined to comprehensively speculate on the motion intention of the interactive obstacle 20, thereby greatly improving the accuracy of speculating on the motion intention of the interactive obstacle 20. For example, when the environmental information of the current vehicle 10 is an intersection and the interactive obstacle 20 is a motor vehicle, the traffic rules for motor vehicles corresponding to this intersection are that it can go straight, turn left or turn right, but cannot make a U-turn. Combining the above information, it can be speculated that the current motion intention of this motor vehicle may be to go straight, turn left or turn right. In addition, the range of its motion intention can be further narrowed by combining the lane where the interactive obstacle 20 is currently located, which will be described in detail below.

[0094] S3: Obtain the passing decision of the current vehicle 10 according to the motion intention.

[0095] Specifically, the motion intention of the interactive obstacle 20 obtained according to the foregoing steps can be used as a consideration factor for the passing decision of the current vehicle 10 to plan the running trajectory of the current vehicle 10.

[0096] In a further embodiment, as Figure 2 shown, before the step S3 of obtaining the passing decision of the current vehicle 10 according to the motion intention, it may further include:

[0097] S01: Obtain the current running trajectory and acceleration of the interactive obstacle 20.

[0098] Specifically, the current running trajectory of the interactive obstacle 20 may be the road 30 where the interactive obstacle 20 is currently located, the lane, the current driving direction, etc.; the acceleration is the current running speed of the interactive obstacle 20 and the rate of change of the running speed, that is, whether the speed of the interactive obstacle 20 is increasing or decreasing during the driving state.

[0099] S02: Obtain the predicted running trajectory of the interactive obstacle 20 according to the current running trajectory, acceleration and motion intention of the interactive obstacle 20.

[0100] Specifically, combine the current running trajectory, acceleration of the interactive obstacle 20, and the motion intention of the interactive obstacle 20 obtained in the foregoing steps to comprehensively judge the predicted running trajectory of the interactive obstacle 20 to judge the possible driving direction, driving speed and running trajectory, etc. of the interactive obstacle 20, so as to provide effective passing decision reference information for the current vehicle 10.

[0101] In an embodiment, as Figure 3 shown, the step S3 of obtaining the passing decision of the current vehicle 10 according to the motion intention may include:

[0102] S31: Select, as the game target 21, the interactive obstacle 20 that has an interactive relationship with the current vehicle 10 from multiple interactive obstacles 20 according to the motion state information and path planning information of the current vehicle 10 at the current moment, and the predicted running trajectory of the interactive obstacle 20.

[0103] Specifically, it can be understood that before obtaining the predicted running trajectory, it is necessary to first obtain the motion state information and path planning information of the current vehicle 10 at the current moment, and obtain the motion state information of the interactive obstacle 20 at the current moment. For example, the motion state information of the current vehicle 10 at the current moment is in motion, and the path planning information is going straight; the motion state information of the interactive obstacle 20 at the current moment is in motion, and the predicted running trajectory of the interactive obstacle 20 may be going straight, turning left or turning right. The interactive obstacle 20 that has an interactive relationship with the current vehicle 10, that is, the predicted running trajectory of this interactive obstacle 20 has overlapping information with the path planning information of the current vehicle 10, then this interactive obstacle 20 can be used as the game target 21 of the current vehicle 10.

[0104] S32: Set an interaction label for the game target 21.

[0105] Specifically, when selecting one or more interactive obstacles 20 as the game target 21, the interaction label of the game target 21 can be further obtained, and the game target 21 can be labeled with the interaction label through the in-vehicle display screen (not shown in the figure) of the current vehicle 10. Among them, the interaction label can include the running direction, right-of-way information, field-of-view information, etc. of the game target 21. The field-of-view information can specifically be whether the current vehicle 10 is in the field of view of the game target 21. A penalty function can also be further set, and safety, passability, right-of-way factors, etc. are mainly considered in the penalty function. In the process of interaction between the interactive obstacle 20 and the current vehicle 10 (autonomous vehicle), the main role of setting the penalty function is to ensure the safety of the current vehicle 10 and optimize the path planning. Specifically, the functions of the penalty function include the following aspects:

[0106] Safe obstacle avoidance:

[0107] The penalty function dynamically adjusts the path planning of the current vehicle 10 by measuring the distance between the current vehicle 10 and the interactive obstacle 20. If the distance between the current vehicle 10 and the obstacle is too close, the value of the penalty function will increase significantly, thereby prompting the current vehicle 10 to take avoidance measures to avoid collisions.

[0108] Path optimization:

[0109] The penalty function is integrated into the objective function to minimize the deviation between the path planning and the global reference path. By adjusting the weight coefficient, the penalty function can avoid obstacles while keeping the current vehicle 10 moving along the path planning information as much as possible, reducing the tracking deviation.

[0110] Dynamic adjustment:

[0111] In a dynamic environment, the positions and states of the interactive obstacles 20 may change continuously. The penalty function can be dynamically adjusted according to the information of the interactively reconstructed obstacles 20 to ensure that the current vehicle 10 is always in a safe area.

[0112] Control constraints:

[0113] The penalty function is combined with the control barrier function to form constraint conditions for the actions of the current vehicle 10. For example, in reinforcement learning, the penalty function can ensure that the current vehicle 10 always meets the safety conditions when exploring the environment, avoiding dangerous behaviors.

[0114] Sparse reward problem:

[0115] In reinforcement learning, the penalty function can also be used to solve the sparse reward problem. By giving penalties or rewards to intermediate states, it helps the current vehicle 10 learn effective obstacle avoidance and passing decisions faster.

[0116] In one embodiment, as Figure 4 shown, the step S32 of setting interactive labels for the game objective 21 may include:

[0117] S321: Sample the accelerations of the current vehicle 10 and the game objective 21 respectively.

[0118] Specifically, first combine Figure 5 to illustrate the vehicle interaction method at the intersection. Figure 5 Shown is an ideal and standard intersection, Figure 5The two interacting objects shown are the current vehicle 10 (ADC) and the vehicle of the interacting obstacle 20 (obstacle). Among them, the four thicker arrowed curves diverging from the front of the current vehicle 10 represent the possible running trajectories of the current vehicle 10, and the four thinner arrowed curves diverging from the front of the interacting obstacle 20 represent the possible running trajectories of the interacting obstacle 20. For the convenience of description, in the following text, ADC_1, ADC_2, ADC_3, and ADC_4 will be uniformly used to represent the path planning information of the ADC, and obstacle_1, obstacle_2, obstacle_3, and obstacle_4 will be used to represent the predicted running trajectories of the interacting obstacle 20. Therefore, when passing through an intersection, there are 4 * 4 = 16 possible combinations of the current vehicle 10 and the interacting obstacle 20. However, this application does not consider the combination method of the paths. Instead, after the path planning information of the current vehicle 10 and the predicted running trajectories of the interacting obstacle 20 are determined, only the acceleration is sampled on their respective running trajectories to generate different longitudinal trajectory clusters 40, and the game is carried out according to the interaction method of the longitudinal trajectory clusters 40. In this application, different lateral path interaction methods are used to assist in annotating the road rights and vision information of the current vehicle 10 and the interacting obstacle 20, and are used to assist in the longitudinal game. For example, when the path of the current vehicle 10_3 and the path of obstacle_4 interact, the current vehicle 10 occupies the main road right, and the interacting obstacle 20 belongs to the behavior of a strong road right. The interacting obstacle 20 is in the vision of the current vehicle 10, and the current vehicle 10 is not in the vision of the interacting obstacle 20. According to this information, the weight parameters of the cost function of the game are dynamically adjusted. The judgment method of whether the interacting obstacle 20 is in its own vision is as follows Figure 6 shown. If the interacting obstacle 20 is within a certain angle range in front of the current vehicle 10, it is considered that the interacting obstacle 20 is in the vision of the current vehicle 10. If the current vehicle 10 is not within a certain angle range (such as 30° - 90°) in front of the interacting obstacle 20, it is considered that the current vehicle 10 is not in the vision of the interacting obstacle 20. For example, it can be judged whether the current vehicle 10 and the interacting obstacle 20 are in each other's vision within 60° on both the left and right sides of a certain position in the vehicle.

[0119] S322: Generate the first sampling space of the current vehicle 10 and the second sampling space of the game target 21 according to the acceleration sampling respectively.

[0120] Specifically, as Figure 7 shown, based on the current motion state information and path planning information of the current vehicle 10 and the current motion state and predicted running trajectories of the interacting obstacle 20, considering longitudinal characteristic information such as cruise speed limit, speed limit of the road 30, path curvature speed limit, and type of the interacting obstacle 20, with the accelerations of the current vehicle 10 and the interacting obstacle 20 as the benchmarks, floating within a certain range (the floating range adopted in this application is plus or minus 1 m / s2 ~2 m / s 2 ), thereby generating a first sampling space of the current vehicle 10 and a second sampling space of the game target 21.

[0121] S323: Deduce the sampling trajectory according to the first sampling space to obtain the first longitudinal trajectory cluster 41 of the current vehicle 10; deduce the sampling trajectory according to the second sampling space to obtain the second longitudinal trajectory cluster 42 of the game target 21.

[0122] Specifically, it can be understood that the longitudinal trajectory cluster 40 includes multiple longitudinal trajectories. During the deduction of the longitudinal trajectory of the current vehicle 10 (ADC), the longitudinal acceleration change of the current vehicle 10 is as Figure 8 shown Figure 8 In the figure, the horizontal axis represents the deduction time t, and the vertical axis represents the acceleration (acc) of the current vehicle 10. The deduction of the longitudinal trajectory of the current vehicle 10 includes the following four aspects.

[0123] (1) Current acceleration maintenance stage (0 = < t < delay_t): In this stage, the longitudinal trajectory of the current vehicle 10 is deduced based on the current acceleration (current_acc) of the current vehicle 10 according to the uniform acceleration model.

[0124] (2) Uniform acceleration change stage (abbreviated as uniform jerk stage) (delay_t = < t < jerk_change_t): Determine the longitudinal acceleration change rate according to the current acceleration of the current vehicle 10 and the sampled target acceleration, and deduce the longitudinal trajectory according to the uniform jerk motion model.

[0125] (3) Uniform acceleration stage (abbreviated as uniform a stage) (jerk_change_t = < t < v_limit_t): In this stage, the current vehicle 10 maintains a longitudinal uniform acceleration for longitudinal trajectory deduction.

[0126] (4) Uniform velocity stage (abbreviated as uniform v stage) (v_limit_t = < t): When the speed reaches the speed limit upper or lower limit during the deduction process, perform longitudinal trajectory deduction on the current vehicle 10 according to the speed limit speed to obtain a first longitudinal trajectory cluster 41 formed by multiple first longitudinal trajectories.

[0127] Based on the longitudinal trajectory deduction of the above four stages and combined with the longitudinal sampling space of the current vehicle 10, 3 longitudinal trajectories can be generated, which is the longitudinal trajectory cluster 40, as Figure 9As shown in the figure. a_curr, a_max, and a_min respectively represent three target accelerations on the sampling space. Among them, the position where a_curr is located is the trajectory generated by the current acceleration, a_curr = current_acc + 0; the position where a_max is located is the trajectory generated by the maximum acceleration, a_max = current_acc + 1.0, and the position where a_min is located is the trajectory generated by the minimum acceleration, a_min = current_acc - 1.0. Figure 9 The three longitudinal trajectories in Figure 9 are the longitudinal trajectories deduced by using the above three accelerations as target accelerations.

[0128] The above four stages of the longitudinal trajectory deduction of the current vehicle 10 may not be fully covered in practice, and it can be determined which stages to adopt according to the actual situation. For example: If the current acceleration (current_acc) of the current vehicle 10 is equal to the longitudinal sampling acceleration (target_acc), then the second stage (the stage of uniform acceleration change) can be omitted.

[0129] The method for deducing the longitudinal trajectory of the interactive obstacle 20 is the same as the method for deducing the longitudinal trajectory of the current vehicle 10 described above, and will not be elaborated here. Similarly, the second longitudinal trajectory cluster 42 can be obtained through this method.

[0130] S324: Perform one-to-one interaction on the first longitudinal trajectory cluster 41 and the second longitudinal trajectory cluster 42 to obtain the interaction cost of each interaction trajectory pair.

[0131] Specifically, the first longitudinal trajectory cluster 41 formed based on the longitudinal deduction trajectory of the current vehicle 10 and the second longitudinal trajectory cluster 42 formed by the longitudinal deduction trajectory of the interactive obstacle 20 can form a 3*3 longitudinal interaction trajectory pair. The optimal longitudinal interaction trajectory pair is selected from these trajectory pairs through a certain standard to represent the current optimal interaction method between the current vehicle 10 and the interactive obstacle 20. In this application, the strategy cost evaluation is performed on all longitudinal interaction trajectory pairs of the current vehicle 10 and the interactive obstacle 20 to obtain the interaction cost (cost). The smaller the strategy cost, the smaller the penalty for this trajectory pair, and the higher the possibility that this strategy is used as the optimal interaction strategy.

[0132] The interaction cost of this application can include multiple sub-cost items. The sub-cost items can be safety cost, comfort cost, efficiency cost, and road right cost. Through the linear combination of these sub-costs, the total cost of each interaction trajectory pair can be calculated.

[0133] In an embodiment, as Figure 10 shown, the step S324 of performing one-to-one interaction on the first longitudinal trajectory cluster 41 and the second longitudinal trajectory cluster 42 to obtain the interaction cost of each interaction trajectory pair may include:

[0134] S3241: Obtain the safety cost for each pair of interaction trajectories.

[0135] Specifically, the safety cost represents a collision penalty. Specifically, collisions can be divided into two collision scenarios:

[0136] Collision Scenario I: When both the current vehicle 10 and the interaction obstacle 20 are in motion, a collision occurs between the side and the tail of the current vehicle 10 and the interaction obstacle 20 in the driving direction of the current vehicle 10, or when the interaction obstacle 20 is stationary, a collision occurs between the current vehicle 10 and any position of the interaction obstacle 20.

[0137] Collision Scenario II: When both the current vehicle 10 and the interaction obstacle 20 are in motion, a collision occurs between the side and the tail of the interaction obstacle 20 and the current vehicle 10 in the driving direction of the interaction obstacle 20, or when the current vehicle 10 is stationary, a collision occurs between the interaction obstacle 20 and any position of the current vehicle 10.

[0138] When designing the safety cost function, if there is no collision between the deduced trajectory of the current vehicle 10 and the deduced trajectory of the interaction obstacle 20 itself, the safety cost is 0; when it is Collision Scenario I, the safety cost is an approximately infinite value; when it is Collision Scenario II, the safety cost is determined according to the speed of the interaction obstacle 20 at the collision position. Specifically, the following calculation method can be used:

[0139]

[0140] where cost safe represents the safety cost cost, v represents the speed of the interaction obstacle 20 at the collision position when the deduced trajectory of the interaction obstacle 20 collides with the deduced trajectory of the current vehicle 10, v min represents the minimum speed for safety cost penalty when the deduced trajectories of the current vehicle 10 and the interaction obstacle 20 collide; v max represents the maximum speed for safety cost penalty when the deduced trajectories of the current vehicle 10 and the interaction obstacle 20 collide, w safe is the weight of the safety cost, and adjusting this value can adjust the intensity of the safety cost penalty.

[0141] S3242: Obtain the comfort cost for each pair of interaction trajectories.

[0142] Specifically, the rate of change of acceleration (i.e., jerk) of a moving object reflects its comfort. The larger the absolute value of jerk, the less comfortable it is; the smaller the absolute value of jerk, the more comfortable it is. Therefore, the comfort cost of the vehicle 10 and the interactive obstacle 20 can be calculated based on the jerk value in the deduced trajectory. Specifically, the comfort cost of the vehicle 10 or the interactive obstacle 20 can be calculated as follows:

[0143]

[0144] Among them, cost comfortable represents the comfort cost, |jerk| represents the absolute value of the acceleration change rate of the current vehicle 10 and the interactive obstacle 20, jerk min Indicates the minimum absolute value of the acceleration change rate between the current vehicle 10 and the interactive obstacle 20, which is generally 0. max is the maximum absolute value of the acceleration change rate of the current vehicle 10 and the interactive obstacle 20, w comfortable The final comfort cost should be a linear combination of the comfort cost of the current vehicle 10 and the comfort cost of the interactive obstacle 20. The comfort penalty weights for the current vehicle 10 and the interactive obstacle 20 can be different and can be set based on different scenarios.

[0145] S3243: Obtain the efficiency cost of each interaction trajectory pair.

[0146] Specifically, the efficiency cost is measured by the time difference between the current vehicle 10 and the game target 21 passing through the path overlap point. The path overlap point is the same path in the predicted running trajectory of the current vehicle 10 and the game target 21, for example Figure 5 The dotted box represents the path overlap point between the current vehicle 10 and the game target 21.

[0147] In a further embodiment, Figure 11 As shown, step S3243 of obtaining the efficiency cost of each interaction trajectory pair may include:

[0148] S32431: Obtain the time when the first longitudinal trajectory cluster 41 and the second longitudinal trajectory cluster 42 pass through the path overlap point respectively.

[0149] Specifically, the time when the first longitudinal trajectory cluster 41 passes the path overlap point can be obtained and recorded as the first time; and the time when the second longitudinal trajectory cluster 42 passes the path overlap point can be obtained and recorded as the second time.

[0150] S32432: Obtain the time it takes for the current vehicle 10 and the game target 21 to pass through the path overlap point at the current acceleration.

[0151] Specifically, that is, obtain the time for the current vehicle 10 to uniformly accelerate through the path coincidence point with the current acceleration, denoted as the third time; obtain the time for the game target 21 to uniformly accelerate through the path coincidence point with the current acceleration, denoted as the fourth time. For example, the times of the third time and the fourth time can both be set within 6 seconds, that is, whether there will be an interaction between the current vehicle 10 and the game target 21 within 6 seconds.

[0152] S32433: Obtain the first difference, as well as the second difference between the maximum value and the minimum value of the first difference.

[0153] Specifically, the first difference can be obtained by taking the difference between the third time and the first time; or the first difference can be obtained by taking the difference between the fourth time and the second time; the maximum value of the first difference is the upper limit of the difference between the third time and the first time, or the upper limit of the difference between the fourth time and the second time; the minimum value of the first difference is the lower limit of the difference between the third time and the first time, or the lower limit of the difference between the fourth time and the second time. The second difference can be obtained by taking the difference between the maximum value and the minimum value of the first difference. For example, the maximum value of the first difference can be 6 - 8 seconds, and the minimum value of the first difference can be 0.

[0154] S32434: Obtain the respective efficiency costs of the current vehicle 10 and the game target 21 through the first difference and the second difference, in combination with the penalty weight of the efficiency cost.

[0155] Specifically, for example, the respective passing efficiencies of the current vehicle 10 and the game target 21 can be obtained first through a normalization method (such as min - max normalization, Min - Max Normalization). Specifically, it can be by taking the difference between the first time or the second time and the minimum value of the above - mentioned first difference, and then obtaining the ratio of this difference to the second difference, so that the data of the current vehicle 10 and the game target 21 are linearly scaled to a specified range (such as [0, 1]). Then, in combination with the penalty weight, adjust the aggressiveness or driving preference of the current vehicle 10. The penalty weights of the current vehicle 10 and the interactive obstacle 20 are different, and the passing efficiency cost consists of the passing efficiency cost of the current vehicle 10 and the passing efficiency cost of the interactive obstacle 20. That is, by assigning different penalty weights to the current vehicle 10 and the interactive obstacle 20, the passing decision is dynamically adjusted. For example, if the penalty weight is large, it tends to a conservative passing decision; if the penalty weight is small, it tends to an aggressive passing decision. Among them, the aggressive passing strategy can be understood as an overtaking strategy, and the conservative passing strategy can be understood as a yielding strategy. In addition, it can also be determined whether the current vehicle 10 adopts an aggressive passing strategy or a conservative passing strategy according to the surrounding environment or the number of interactive obstacles 20. The penalty weight coefficient can be specifically identified according to different scenarios. This application mainly focuses on the intersection scenario.

[0156] Furthermore, the calculation method of the passing efficiency cost function of the current vehicle 10 or the interactive obstacle 20 can be as follows:

[0157]

[0158] where cost efficiency represents the passing efficiency cost, delta_t represents the difference between the time when the longitudinal deduction trajectory of the current vehicle 10 or the interactive obstacle 20 passes through the path coincidence point and the time when the current vehicle 10 or the interactive obstacle 20 uniformly accelerates through the path coincidence point with the current acceleration, t min represents the lower limit of the time difference, t max represents the upper limit of the time difference, w efficiency represents the penalty weight of the passing efficiency cost.

[0159] S3244: Obtain the road right cost of each interactive trajectory pair.

[0160] Specifically, the road right represents the level of the driving right of the current vehicle 10 and the interactive obstacle 20 for the current road 30. According to the interactive trajectory pair of the current vehicle 10 and the interactive obstacle 20, judge the road right relationship between the current vehicle 10 and the interactive obstacle 20. If the game causes the vehicle with the higher road right to change its motion state, then the change of this motion state should be punished with a higher road right cost, that is, the vehicle with the higher road right in motion tends not to change its current motion state. The specific road right cost of the interactive trajectory pair of the current vehicle 10 or the interactive obstacle 20 is calculated as follows:

[0161]

[0162] where cost roadright represents the road right cost of the current moving vehicle (which can represent the current vehicle 10 or the interactive obstacle 20), a represents the current longitudinal sampled acceleration of the current vehicle 10 or the interactive obstacle 20, a max represents the upper limit of the longitudinal acceleration sampling, a min represents the lower limit of the longitudinal acceleration sampling, w roadright represents the penalty weight of the road right cost, which is adjusted according to the current moving vehicle and the road right information. The road right cost is composed of the road right costs of the current vehicle 10 and the interactive obstacle 20 respectively, and w is adjusted by comprehensively considering the road right situation and the vision situation it occupies roadrightThe size. For example, the penalty weights for the current vehicle 10 and the interaction obstacle 20 that do not have the right of way are high. For example, in the scenario of highway merging, if the current vehicle 10 merges from the side road into the main road, then the current vehicle 10 does not have the right of way, and the overtaking space of the current vehicle 10 is calculated to determine whether to yield or not. If the overtaking space is sufficient, an overtaking strategy is adopted; if the overtaking space is insufficient, a yielding strategy is adopted. The situation considering the field of view can be: when the interaction obstacle 20 is within the field of view of the current vehicle 10 (the vehicle itself) and the current vehicle 10 is not within the field of view of the interaction obstacle 20, a conservative passing strategy, that is, a yielding strategy, is preferred; when the current vehicle 10 is within the field of view of the interaction obstacle 20 and the interaction obstacle 20 is not within the field of view of the current vehicle 10, it is considered that the interaction obstacle 20 will respond to the acceleration or deceleration of the current vehicle 10. Therefore, at this time, the current vehicle 10 can tend to an aggressive passing strategy, that is, an overtaking strategy, that is, the penalty weight of the current vehicle 10 is low.

[0163] S3245: By summing up the safety cost, comfort cost, efficiency cost, and right-of-way cost, the total cost of the interaction between the current vehicle 10 and the game target 21 is obtained.

[0164] Specifically, the total cost (cost total ) can be calculated by adding up all the above sub-cost items, as follows:

[0165] cost total = cost safe + cost comfortable + cost efficiency + cost roadright

[0166] S325: Select the interaction trajectory pair with the lowest interaction cost from the interaction costs of the interaction trajectory pairs as the interaction method with the highest running probability between the current vehicle 10 and the game target 21.

[0167] Specifically, the trajectory pair with the minimum total cost is selected from the interaction trajectory pairs of the current vehicle 10 and the interaction obstacle 20 as the optimal interaction method, that is, the interaction method with the highest running probability, and the overtaking strategy or yielding strategy of the current vehicle 10 with respect to the interaction obstacle 20 is determined according to the time difference between the current vehicle 10 and the interaction obstacle 20 reaching the interaction position.

[0168] S33: According to the interaction label, obtain the interaction method with the highest running probability between the current vehicle 10 and the game target 21 to obtain the passing decision of the current vehicle 10.

[0169] Specifically, such as Figure 3 and Figure 5As shown in the figure, according to the above-mentioned interaction tags and penalty function set for the game target 21, the interaction method with the highest running probability between the current vehicle 10 and the game target 21 is obtained. The interaction method with the highest probability can be understood as permuting and combining the trajectories of the sampled interaction obstacle 20 and the current vehicle 10, and screening out the interaction method with the smallest penalty function cost, that is, the optimal interaction method, so as to determine the passing decision of the current vehicle 10 for the interaction obstacle 20. The passing decision can include an overtaking game strategy, a yielding game strategy, etc.

[0170] In one embodiment, as Figure 12 shown in the figure, the step S33 of obtaining the passing decision of the current vehicle 10 by obtaining the interaction method with the highest running probability between the current vehicle 10 and the game target 21 according to the interaction tag may include:

[0171] S331: In response to the current vehicle 10 arriving at the path coincidence point earlier than the game target 21, the passing decision of the current vehicle 10 is an overtaking game strategy.

[0172] S332: In response to the current vehicle 10 arriving at the path coincidence point later than the game target 21, the passing decision of the current vehicle 10 is a yielding game strategy.

[0173] Specifically, if the current vehicle 10 arrives at the interaction position earlier than the interaction obstacle 20, it is considered that the current vehicle 10 can safely pass through the interaction position before the interaction obstacle 20 arrives at the interaction position. Therefore, at this time, the current vehicle 10 overtakes relative to the interaction obstacle 20, that is, adopts an overtaking game strategy; if the current vehicle 10 arrives at the interaction position later than the interaction obstacle 20, it is considered that the current vehicle 10 cannot safely pass through the interaction position before the interaction obstacle 20 arrives at the interaction position. Therefore, at this time, the current vehicle 10 yields relative to the interaction obstacle 20, that is, a yielding game strategy.

[0174] In another embodiment, as Figure 13 shown in the figure, after the step S3 of obtaining the passing decision of the current vehicle 10 according to the motion intention, it may include:

[0175] S4: According to the interaction method with the highest running probability between the current vehicle 10 and the game target 21, and according to the passing decision, generate the running trajectory information of the current vehicle 10.

[0176] Specifically, the deduced trajectory of the interaction obstacle 20 in the optimal interaction trajectory pair obtained in the above steps can be used to replace the predicted trajectory of the current interaction obstacle 20 for the use of the downstream planning module 150. According to the passing decision (such as overtaking game strategy or yielding game strategy), guide the downstream to perform speed planning to generate safe, comfortable and efficient running trajectory information for the current vehicle 10.

[0177] In another embodiment, the traffic planning method provided by the present application may further include:

[0178] S5: Obtain the operation control information of the current vehicle 10 according to the operation trajectory information of the current vehicle 10.

[0179] Specifically, after obtaining the operation trajectory information of the current vehicle 10, the control module 140 may send control information to the execution end of the current vehicle 10, so that the current vehicle 10 can operate according to the operation trajectory information.

[0180] S6: Send the operation control information to the execution end of the current vehicle 10.

[0181] Specifically, the operation control information includes the operation acceleration of the current vehicle 10 and the direction control information. After the execution end of the current vehicle 10 obtains the operation control information sent by the control module 140, it can control the current vehicle 10 to travel according to the corresponding operation acceleration and direction according to the operation control information. For example, if the operation acceleration in the operation control information is to accelerate to a certain speed at the current moment and the direction control information is to go straight, the execution end of the current vehicle 10 controls the current vehicle 10 to travel according to the operation control information.

[0182] It can be understood that the specific contents of steps S1 to S3 in this embodiment are the same as those in the foregoing embodiment, and will not be elaborated here.

[0183] To solve the above technical problems, the present application also provides a traffic planning system 100.

[0184] As Figure 14 shown, the traffic planning system 100 provided by an embodiment of the present application may include: a perception and prediction module 110, a judgment module 120, a decision-making module 130, and a control module 140. The perception and prediction module 110 is used to obtain the type characteristics of the interactive obstacles 20 around the current vehicle 10; the perception and prediction module 110 mainly relies on the hardware sensors of the current vehicle 10 (not shown in the figure) to sense the surrounding environment information and uses a prediction algorithm to generate the prediction trajectory of the interactive obstacles 20. The judgment module 120 is used to infer the movement intention of the interactive obstacles 20 according to the type characteristics of the interactive obstacles 20. The decision-making module 130 is used to obtain the traffic decision of the current vehicle 10 according to the movement intention, for example, decide traffic decisions such as overtaking and giving way for the current vehicle 10 next.

[0185] As Figure 15As shown, in a further embodiment, the traffic planning system 100 may further include a planning module 150. The planning module 150 generates reasonable motion trajectory information of the current vehicle 10 based on the information generated by the decision module 130 and the information of the interactive obstacle 20 output by the perception and prediction module 110. The control module 140 is connected to the perception and prediction module 110, the judgment module 120, and the decision module 130, and is used to obtain the information of the corresponding modules and send control instructions to the corresponding modules. For example, the control module 140 receives the motion trajectory information of the current vehicle 10 output by the upstream planning module 150, and then uses a preset control algorithm to generate acceleration, direction information, etc. that enable the current vehicle 10 to travel along the planned trajectory. The traffic planning system 100 may further include a vehicle actuator module 170. The vehicle actuator module 170 may be the execution end of the vehicle and can respond in real time to the control instructions of the current vehicle 10 sent by the control module 140 to further control the specific driving information of the current vehicle 10.

[0186] As Figure 15 and Figure 16 As shown, in a further embodiment, the traffic planning system 100 may further include a game target screening module 160. First, by receiving the motion state information and path planning information of the current vehicle 10 at the current moment, and the motion state information and predicted trajectory information for a future period of time of the interactive obstacle 20, the game target screening module 160 can screen out the main game target 21 from the interactive obstacles 20 that have an interactive relationship with the motion trajectory of the current vehicle 10, and label the selected main game target 21 with an interactive label. The interactive label includes going straight or turning, whether to occupy the right of way, whether the current vehicle 10 is in the field of view, etc.

[0187] Secondly, the decision module 130 may specifically be an interactive game decision module. The interactive game decision module receives the main game target 21 output by the game target screening module 160, and combines the going straight / turning information, right of way information, and field of view information of the current vehicle 10 and the interactive obstacle 20, respectively generates a sampling space and deduces sampling trajectories for the current vehicle 10 and the interactive obstacle 20, then pairwise interacts the longitudinal trajectory clusters 40 deduced for the current vehicle 10 and the interactive obstacle 20, calculates the cost of each interactive trajectory pair, selects the interactive trajectory pair with the lowest cost, and generates a traffic decision label for the current vehicle 10 according to this interaction method, such as overtaking, yielding, etc. decision labels. Finally, after receiving the overtaking and yielding decisions of the current vehicle 10 for the main interactive obstacle 20, the planning module 150 will generate a safe and comfortable motion trajectory of the current vehicle 10 through a speed optimizer according to this decision.

[0188] Regarding the main working methods and processes of the traffic planning system 100, please refer to the foregoing traffic planning method, which will not be elaborated herein.

[0189] The traffic planning method provided by this application includes: obtaining the type characteristics of interactive obstacles around the current vehicle; inferring the motion intention of the interactive obstacles according to the type characteristics of the interactive obstacles; and obtaining the traffic decision of the current vehicle according to the motion intention. This application infers the motion intention of dynamic obstacles according to their type characteristics, so as to provide decision-making information for the current vehicle, solve the problems of low traffic efficiency, low safety and poor experience of autonomous vehicles at intersections caused by frequent point braking, awkward stopping and other problems of autonomous vehicles, with strong flexibility, high traffic efficiency and high safety, while improving the experience.

[0190] Please refer to Figure 17 , Figure 17 which is a schematic structural diagram of an electronic device provided by an embodiment of this application.

[0191] The electronic device 200 may specifically include a processor 210 and a memory 220. The memory 220 is coupled to the processor 210.

[0192] The processor 210 is used to control the operation of the electronic device 200. The processor 210 may also be referred to as a CPU (Central Processing Unit, central processing unit). The processor 210 may be an integrated circuit chip with signal processing capabilities. The processor 210 may also be a general-purpose processor, 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, discrete hardware components. The general-purpose processor may be a microprocessor or the processor 210 may also be any conventional processor, etc.

[0193] The memory 220 is used to store computer programs, which may be RAM, ROM, or other types of storage devices. Specifically, the memory may include one or more computer-readable storage media, and the computer-readable storage media may be non-transitory. The memory may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices and flash storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory is used to store at least one program code.

[0194] The processor 210 is used to execute the computer program stored in the memory 220 to implement the traffic planning method described in the embodiment of the traffic planning method of this application.

[0195] In some embodiments, the electronic device 200 may further include: a peripheral device interface 230 and at least one peripheral device. The processor 210, the memory 220, and the peripheral device interface 230 may be connected through a bus or signal lines. Each peripheral device may be connected to the peripheral device interface 230 through a bus, signal lines, or a circuit board. Specifically, the peripheral device includes at least one of a radio frequency circuit 240, a display screen 250, an audio circuit 260, and a power supply 270.

[0196] The peripheral device interface 230 can be used to connect at least one peripheral device related to I / O (Input / Output) to the processor 210 and the memory 220. In some embodiments, the processor 210, the memory 220, and the peripheral device interface 230 are integrated on the same chip or circuit board; in some other embodiments, any one or two of the processor 210, the memory 220, and the peripheral device interface 230 can be implemented on a separate chip or circuit board, and this embodiment does not limit this.

[0197] The radio frequency circuit 240 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The radio frequency circuit 240 communicates with a communication network and other communication devices through electromagnetic signals, and the radio frequency circuit 240 is the communication circuit of the electronic device 200. The radio frequency circuit 240 converts an electrical signal into an electromagnetic signal for transmission, or converts the received electromagnetic signal into an electrical signal. Optionally, the radio frequency circuit 240 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a subscriber identity module card, and so on. The radio frequency circuit 240 can communicate with other terminals through at least one wireless communication protocol. The wireless communication protocol includes but is not limited to: the World Wide Web, a metropolitan area network, an intranet, each generation of mobile communication networks (2G, 3G, 4G, and 5G), a wireless local area network, and / or a WiFi (Wireless Fidelity) network. In some embodiments, the radio frequency circuit 240 may further include a circuit related to NFC (Near Field Communication), and this application does not limit this.

[0198] The display screen 250 is used to display the UI (User Interface). The UI may include graphics, text, icons, videos, and any combination thereof. When the display screen 250 is a touch display screen, the display screen 250 also has the ability to collect touch signals on or above the surface of the display screen 250. The touch signals can be input into the processor 210 as control signals for processing. At this time, the display screen 250 can also be used to provide virtual buttons and / or virtual keyboards, also known as soft buttons and / or soft keyboards. In some embodiments, there may be one display screen 250, which is disposed on the front panel of the electronic device 200; in other embodiments, there may be at least two display screens 250, which are respectively disposed on different surfaces of the electronic device 200 or in a folded design; in other embodiments, the display screen 250 may be a flexible display screen, which is disposed on the curved surface or the folding surface of the electronic device 200. Even, the display screen 250 can also be set to an irregular non-rectangular shape, that is, a special-shaped screen. The display screen 250 can be prepared from materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0199] The audio circuit 260 may include a microphone and a speaker. The microphone is used to collect sound waves of the usage and the environment, and convert the sound waves into electrical signals and input them into the processor 210 for processing, or input them into the radio frequency circuit 240 to achieve voice communication. For the purpose of stereo collection or noise reduction, there may be multiple microphones, which are respectively disposed at different parts of the electronic device 200. The microphone can also be an array microphone or an omnidirectional collection microphone. The speaker is used to convert the electrical signals from the processor 210 or the radio frequency circuit 240 into sound waves. The speaker can be a traditional thin film speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can not only convert the electrical signals into audible sound waves, but also convert the electrical signals into inaudible sound waves for uses such as ranging. In some embodiments, the audio circuit 260 may further include a headphone jack.

[0200] The power supply 270 is used to supply power to each component in the electronic device 200. The power supply 270 can be alternating current, direct current, a disposable battery, or a rechargeable battery. When the power supply 270 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery charged through a wired line, and a wireless rechargeable battery is a battery charged through a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0201] For a detailed description of the functions and execution processes of each functional module or component in the embodiment of the electronic device 200 of the present application, reference may be made to the description in the embodiment of the general access planning method of the present application above, and details will not be repeated here.

[0202] In several embodiments provided by the present application, it should be understood that the disclosed electronic device 200 and access planning method can be implemented in other ways. For example, the embodiments of the electronic device 200 described above are merely illustrative. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other can be through some interfaces, and the indirect coupling or communication connection of devices or units can be in electrical, mechanical or other forms.

[0203] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0204] In addition, in each embodiment of the present application, the functional units can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0205] Please refer to Figure 18 , Figure 18 is a structural schematic diagram of a computer-readable storage medium provided by an embodiment of the present application.

[0206] Refer to Figure 18, when the above integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in the computer-readable storage medium 300. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions / computer programs to enable an electronic device (which can be a computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, as well as electronic devices such as computers, mobile phones, laptop computers, tablet computers, cameras, etc. having the above storage media.

[0207] The description of the execution process of the program data in the computer-readable storage medium 300 can be referred to the embodiments of the general planning method of the present application described above, and will not be repeated here.

[0208] The above is only the implementation mode of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A traffic planning method, characterized in that, Including: Obtain the type characteristics of the interactive obstacles (20) around the current vehicle (10); Infer the movement intention of the interactive obstacles (20) according to the type characteristics of the interactive obstacles (20); Obtain the passing decision of the current vehicle (10) according to the movement intention.

2. The method according to claim 1, wherein Before inferring the movement intention of the interactive obstacles (20) according to the type characteristics of the interactive obstacles (20), it further includes: Obtain the environmental information of the current vehicle (10); The inferring the movement intention of the interactive obstacles (20) according to the type characteristics of the interactive obstacles (20) includes: Infer the movement intention of the interactive obstacles (20) according to the type characteristics of the interactive obstacles (20) themselves, the traffic rules corresponding to the type characteristics, and the environmental information.

3. The method according to claim 1, characterized in that Before obtaining the passing decision of the current vehicle (10) according to the movement intention, it further includes: Obtain the current running trajectory and acceleration of the interactive obstacles (20); Obtain the predicted running trajectory of the interactive obstacles (20) according to the current running trajectory, the acceleration, and the movement intention of the interactive obstacles (20).

4. The method according to claim 3, wherein The obtaining the passing decision of the current vehicle (10) according to the movement intention includes: According to the movement state information and path planning information of the current vehicle (10) at the current moment, and the predicted running trajectory of the interactive obstacles (20), select the interactive obstacles (20) having an interactive relationship with the current vehicle (10) from the multiple interactive obstacles (20) as the game target (21); Set an interaction label for the game target (21); wherein the interaction label includes the running direction, road right information, and field of view information of the game target (21); According to the interaction label, obtain the interaction method with the highest running probability between the current vehicle (10) and the game target (21), so as to obtain the passing decision of the current vehicle (10).

5. The method according to claim 4, wherein The setting the interaction label for the game target (21) includes: Perform acceleration sampling on the current vehicle (10) and the game target (21) respectively; Generate a first sampling space of the current vehicle (10) and a second sampling space of the game target (21) respectively according to the acceleration sampling; Deduce the sampling trajectory according to the first sampling space to obtain the first longitudinal trajectory cluster (41) of the current vehicle (10); deduce the sampling trajectory according to the second sampling space to obtain the second longitudinal trajectory cluster (42) of the game target (21); Perform one-to-one interaction on the first longitudinal trajectory cluster (41) and the second longitudinal trajectory cluster (42) to obtain the interaction cost of each interaction trajectory pair Select the interaction trajectory pair with the lowest interaction cost from the interaction costs of the interaction trajectory pairs as the interaction method with the highest running probability between the current vehicle (10) and the game target (21).

6. The method according to claim 5, wherein the interaction cost includes a safety cost, a comfort cost, an efficiency cost, and a right-of-way cost; the efficiency cost is measured by the time difference between the current vehicle (10) and the game target (21) passing through the path coincidence point; wherein, the path coincidence point is the same path in the predicted running trajectories of the current vehicle (10) and the game target (21); Performing one-to-one interaction on the first longitudinal trajectory cluster (41) and the second longitudinal trajectory cluster (42) to obtain the interaction cost of each interaction trajectory pair includes: Obtaining the safety cost of each interaction trajectory pair; Obtaining the comfort cost of each interaction trajectory pair; Obtaining the efficiency cost of each interaction trajectory pair; Obtaining the right-of-way cost of each interaction trajectory pair; By summing up the safety cost, the comfort cost, the efficiency cost, and the right-of-way cost, obtaining the total cost of the interaction between the current vehicle (10) and the game target (21).

7. The method according to claim 6, wherein obtaining the efficiency cost of each interaction trajectory pair includes: Obtaining the times when the first longitudinal trajectory cluster (41) and the second longitudinal trajectory cluster (42) respectively pass through the path coincidence point; Obtaining the times when the current vehicle (10) and the game target (21) respectively pass through the path coincidence point with uniform acceleration at the current acceleration; Obtaining a first difference, and a second difference between the maximum value and the minimum value of the first difference; Based on the first difference and the second difference, and in combination with the penalty weight of the efficiency cost, obtaining the respective efficiency costs of the current vehicle (10) and the game target (21).

8. The method according to claim 7, wherein obtaining the interaction method with the highest running probability between the current vehicle (10) and the game target (21) according to the interaction label to obtain the passing decision of the current vehicle (10) includes: In response to the current vehicle (10) arriving at the path coincidence point earlier than the game target (21), the passing decision of the current vehicle (10) is an overtaking game strategy; In response to the current vehicle (10) arriving at the path coincidence point later than the game target (21), the passing decision of the current vehicle (10) is a yielding game strategy.

9. The method according to any one of claims 4 to 8, wherein after obtaining the passing decision of the current vehicle (10) according to the motion intention, it includes: Generating the running trajectory information of the current vehicle (10) according to the interaction method with the highest running probability between the current vehicle (10) and the game target (21), and according to the passing decision.

10. The method according to claim 9, wherein It further includes: Obtain the operation control information of the current vehicle (10) according to the operation trajectory information of the current vehicle (10); Send the operation control information to the execution end of the current vehicle (10); wherein the operation control information includes the operation acceleration and direction control information of the current vehicle (10).

11. A passage planning system (100), characterized in that, Comprising: A perception and prediction module (110) for obtaining the type characteristics of the interactive obstacles (20) around the current vehicle (10); A judgment module (120) for inferring the movement intention of the interactive obstacle (20) according to the type characteristics of the interactive obstacle (20); A decision-making module (130) for obtaining the passage decision of the current vehicle (10) according to the movement intention; A control module (140), connected to the perception and prediction module (110), the judgment module (120) and the decision-making module (130), for obtaining the information of the corresponding module and sending control instructions to the corresponding module.

12. An electronic device (200), characterized in that, Comprising: A processor (210); A memory (220), connected to the processor (210), for storing a computer program that can run on the processor (210); Wherein, when the processor (210) executes the computer program, the method according to any one of claims 1 to 10 is implemented.

13. A computer-readable storage medium (300), characterized in that, A computer program is stored on a computer-readable storage medium (300), and when the computer program is executed by a processor (210), the method according to any one of claims 1 to 10 is implemented.

Images