Vehicle meeting trajectory prediction method and device, electronic equipment, vehicle and product

Abstract

The invention provides a vehicle meeting track prediction method and device, electronic equipment, a vehicle and a product, and the method comprises the steps: responding to a detected opposite coming vehicle, carrying out the traversal of a behavior sample in a pre-constructed driving normal form behavior space, and enabling the behavior sample to comprise the longitudinal acceleration and transverse deviation parameters of the vehicle; after one behavior sample is traversed every time, track deduction is carried out on the self-vehicle and the opposite-direction coming vehicle according to the behavior sample and current motion parameters of the self-vehicle and the opposite-direction coming vehicle, and track pairs of the self-vehicle and the opposite-direction coming vehicle are obtained; carrying out quantitative evaluation on a plurality of track pairs obtained after traversal is finished so as to determine a target track pair adaptive to the own vehicle; and controlling the self-vehicle to run according to the self-vehicle deduction trajectory in the target trajectory pair. Accordingly, traversal and trajectory deduction are carried out on the driving normal form behavior space, trajectory prediction during vehicle meeting is achieved, and safety and passing efficiency during vehicle meeting of the automatic driving vehicles are improved.

Description

Technical Field

[0001] One or more embodiments of the present disclosure relate to the technical field of trajectory prediction, and in particular, to a method, apparatus, electronic device, and storage medium for predicting oncoming vehicle trajectories. Background Art

[0002] With the rapid development of autonomous driving technology, trajectory prediction and decision-making planning of vehicles in complex traffic scenarios have become a research hotspot. Especially in oncoming vehicle scenarios, due to the complex interaction and game relationship between vehicles, how to accurately predict vehicle trajectories and make reasonable decisions is of great significance for ensuring driving safety.

[0003] Currently, trajectory prediction in oncoming vehicle scenarios requires independently predicting the future trajectory of one vehicle first and then planning a path for the other vehicle based on it. For example, first, separately predict the movement trajectory of the oncoming vehicle, and then generate a feasible path for the host vehicle based on the space occupied by its trajectory. Such methods separate the decision-making processes of the host vehicle and the oncoming vehicle, ignoring the dynamic game relationship in which the behaviors of both parties affect and co-evolve with each other during the actual oncoming vehicle process, and it is difficult to reflect the interaction characteristics in real traffic scenarios. Therefore, the generated decisions are often limited to local optimality, lacking global coordination, and are prone to decision failure when facing unexpected or complex vehicle behaviors, thereby affecting driving safety and traffic efficiency. Summary of the Invention

[0004] To solve the problem in the related art that the trajectory prediction method does not consider the mutual influence of the behaviors of the host vehicle and the oncoming vehicle during the oncoming vehicle process, resulting in the lack of global coordination of the generated prediction trajectories, the present disclosure provides a method for predicting oncoming vehicle trajectories, and the method includes: In response to detecting an oncoming vehicle, traverse the behavior samples in the pre-constructed driving paradigm behavior space, where the behavior samples include the longitudinal acceleration and lateral offset parameters of the vehicle; After traversing each behavior sample, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, to obtain a trajectory pair of the host vehicle and the oncoming vehicle; Quantitatively evaluate the several trajectory pairs obtained after the traversal ends, to determine the target trajectory pair that suits the host vehicle; Control the host vehicle to drive according to the deduced trajectory of the host vehicle in the target trajectory pair.

[0005] Optionally, before traversing the behavior samples in the pre-constructed driving paradigm behavior space, the method further includes: Obtain a vehicle real driving data set, where each data in the vehicle real driving data set includes the longitudinal acceleration and lateral offset parameters of vehicle oncoming. Filter the data in the vehicle's real driving dataset, only retain the representative data of different driving modes, and construct a driving paradigm behavior space based on the representative data of different driving modes.

[0006] Optionally, after traversing to a behavior sample each time, based on the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively to obtain a trajectory pair of the host vehicle and the oncoming vehicle, including: After traversing to a behavior sample each time, based on the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively for a period of time corresponding to one deduction after the current moment; Re-traverse the behavior samples in the driving paradigm behavior space. When traversing to another behavior sample, repeat the above trajectory deduction process until the deduced trajectory pair is sufficient for the host vehicle and the oncoming vehicle to end the passing.

[0007] Optionally, the method further includes: Estimate the time required for the host vehicle and the oncoming vehicle to end the passing, so as to determine at least the preset number of deduction processes according to the time corresponding to one deduction and the time required to end the passing; If the deduced trajectory pair has completed the preset number of deduction processes, it is determined that the deduced trajectory pair is sufficient for the host vehicle and the oncoming vehicle to end the passing.

[0008] Optionally, after controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: If the passing process between the host vehicle and the oncoming vehicle has not been completed, predict the ideal position of the oncoming vehicle according to the deduced trajectory of the oncoming vehicle in the target trajectory pair; Detect the deviation between the actual position and the ideal position of the oncoming vehicle. If the deviation exceeds the preset deviation threshold, re-traverse the behavior samples in the pre-constructed driving paradigm behavior space to determine the target trajectory pair suitable for the host vehicle again; If the deviation does not exceed the preset deviation threshold, continue to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

[0009] Optionally, controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair includes: controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair and accumulating the number of decision steps at one time; After controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: If the passing process between the host vehicle and the oncoming vehicle is not completed, the cumulative number of decision steps is counted; If the cumulative number of decision steps reaches a preset step threshold, the behavior samples in the pre-constructed driving paradigm behavior space are traversed again to determine the target trajectory pair suitable for the host vehicle again; If the cumulative number of decision steps does not reach the preset step threshold, the host vehicle is continuously controlled to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, and the number of decision steps is cumulatively counted once.

[0010] Optionally, before quantitatively evaluating the several trajectory pairs obtained after the traversal to determine the target trajectory pair suitable for the host vehicle, the method further includes: Performing a collision risk assessment on the several trajectory pairs obtained after the traversal according to a preset risk assessment rule; According to the collision risk assessment result, only retain the several trajectory pairs with qualified collision risk assessment results.

[0011] The present disclosure also provides a passing trajectory prediction device, and the device includes: A traversal unit, configured to traverse the behavior samples in the pre-constructed driving paradigm behavior space in response to detecting an oncoming vehicle, where the behavior samples include the longitudinal acceleration and lateral offset parameters of the vehicle; A deduction unit, configured to, after traversing each behavior sample, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, to obtain a trajectory pair of the host vehicle and the oncoming vehicle; An evaluation unit, configured to quantitatively evaluate the several trajectory pairs obtained after the traversal to determine the target trajectory pair suitable for the host vehicle; A control unit, configured to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

[0012] The present disclosure also provides a vehicle, including: A processor; A memory for storing instructions executable by the processor; Wherein, the processor realizes the above method by running the executable instructions.

[0013] The present disclosure also provides a computer program product, including a computer program / instructions, and when the computer program / instructions are executed by a processor, the above method is realized.

[0014] According to embodiments of the present disclosure, in response to detecting an oncoming vehicle, behavior samples in a pre-constructed driving paradigm behavior space including vehicle longitudinal acceleration and lateral offset parameters are traversed. For each behavior sample, trajectory deduction is performed based on the motion parameters of the current host vehicle and the oncoming vehicle to generate a series of possible trajectory pairs of the host vehicle and the oncoming vehicle. Subsequently, these trajectory pairs are quantitatively evaluated, and the target trajectory pairs suitable for the host vehicle are selected therefrom as the final decision basis. Finally, the host vehicle is controlled to drive according to the host vehicle trajectory in the selected target trajectory pair.

[0015] In the above manner, compared with the defect that traditional step-by-step decision-making cannot achieve global optimality, the technical solution of the present disclosure regards the host vehicle and the oncoming vehicle as equal game players, can dynamically balance the behaviors of each vehicle, achieve more reasonable decision-making planning in complex oncoming vehicle scenarios, avoid the decision-making limitations brought by unilateral prediction, significantly improve the accuracy and reliability of oncoming vehicle trajectory prediction, enhance the system's ability to cope with dynamic environmental changes, and effectively ensure the safety and efficiency of driving. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments recorded in the present disclosure. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 is a flowchart of a method for predicting an oncoming vehicle trajectory shown in an exemplary embodiment; Figure 2 is a schematic diagram of vehicle trajectory deduction in an oncoming vehicle scenario shown in an exemplary embodiment; Figure 3 is a flowchart of another method for predicting an oncoming vehicle trajectory shown in an exemplary embodiment; Figure 4 is a hardware structure diagram of an electronic device shown in an exemplary embodiment; Figure 5 is a block diagram of a device for predicting an oncoming vehicle trajectory shown in an exemplary embodiment. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To enable those skilled in the art to better understand the technical solutions in the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present disclosure. Obviously, the described embodiments are only a part of the embodiments of the present disclosure, rather than all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present disclosure.

[0019] It should be noted that: in other embodiments, the steps of the corresponding methods are not necessarily executed in the order shown and described in the present disclosure. In some other embodiments, the steps included in the method may be more or less than those described in the present disclosure. In addition, a single step described in the present disclosure may be decomposed into multiple steps for description in other embodiments; and multiple steps described in the present disclosure may also be combined into a single step for description in other embodiments.

[0020] With the rapid development of autonomous driving technology, trajectory prediction and decision-making planning of vehicles in complex traffic scenarios have become a research hotspot. Especially in the oncoming vehicle scenario, due to the complex interaction and game relationship between vehicles, how to accurately predict vehicle trajectories and make reasonable decisions is of great significance for ensuring driving safety.

[0021] Currently, trajectory prediction in the oncoming vehicle scenario requires first independently predicting the future trajectory of one vehicle and then planning a path for the other vehicle based on this. For example, first separately predict the motion trajectory of the oncoming vehicle, and then generate a feasible path for the host vehicle based on the space occupied by its trajectory. Such methods separate the decision-making processes of the host vehicle and the oncoming vehicle, ignoring the dynamic game relationship in which the behaviors of both sides affect and co-evolve with each other during the actual oncoming vehicle process, and it is difficult to reflect the interaction characteristics in the real traffic scenario. Therefore, the generated decisions are often limited to local optimality, lacking global coordination, and are prone to decision failure when facing unexpected or complex vehicle behaviors, thus affecting driving safety and traffic efficiency.

[0022] In view of this, the present disclosure aims to propose a technical solution for oncoming vehicle trajectory prediction that fully considers the interaction and game relationship between vehicles.

[0023] This technical solution first responds to detecting an oncoming vehicle, and traverses the behavior samples in the pre-constructed driving paradigm behavior space that includes vehicle longitudinal acceleration and lateral offset parameters; further, for each behavior sample, based on the motion parameters of the current host vehicle and the oncoming vehicle, trajectory deduction is performed to generate a series of possible trajectory pairs of the host vehicle and the oncoming vehicle. Subsequently, these trajectory pairs are quantitatively evaluated, and the target trajectory pairs that are suitable for the host vehicle are selected therefrom as the final decision-making basis. Finally, the host vehicle is controlled to drive according to the host vehicle trajectory in the selected target trajectory pair.

[0024] For example, vehicle A is traveling at a speed of 60 km / h on a two-lane, two-way road. At this time, it is detected that a oncoming vehicle B is approaching at a speed of 70 km / h, and the distance between the two vehicles is 100 m. The system of vehicle A triggers the oncoming vehicle trajectory prediction method and traverses the behavior samples from the pre-constructed driving paradigm behavior space. Suppose the following behavior samples are included in this behavior space: Sample 1: longitudinal acceleration = 0 m / s² (maintaining a constant speed), lateral offset = 0.2 m (offset 0.2 m to the right side of the lane); Sample 2: longitudinal acceleration = 0.1 m / s² (accelerating), lateral offset = -0.1 m (offset 0.1 m to the left side of the lane); Sample 3: longitudinal acceleration = -0.3 m / s² (decelerating), lateral offset = -0.3 m (offset 0.3 m to the left side of the lane). After traversing to Sample 1, the trajectories of vehicle A and vehicle B are deduced based on Sample 1 combined with the current motion parameters of vehicle A and vehicle B. It can be known that next, vehicle A will maintain a constant speed of 60 km / h and at the same time offset 0.2 m to the right side of the lane; vehicle B will maintain a constant speed of 70 km / h, and at the same time the lateral offset of vehicle B is 0 (continuing to drive at the original lateral position of the lane). According to the above trajectory deduction process, Trajectory Pair 1 of vehicle A and vehicle B is obtained. After traversing to Sample 2 and Sample 3, in a similar process as above, Trajectory Pair 2 and Trajectory Pair 3 of vehicle A and vehicle B are obtained. Then, the system quantitatively evaluates the 3 groups of trajectory pairs obtained after the traversal from indicators such as safety, comfort, and traffic efficiency. After evaluation, the trajectory pair of Sample 1 has the highest score and is determined as the target trajectory pair suitable for vehicle A. Finally, the system controls vehicle A to drive according to the deduced trajectory of the ego vehicle in the target trajectory pair, that is, vehicle A maintains a constant speed of 60 km / h and at the same time offsets 0.2 m to the right side of the lane, thereby safely, comfortably and efficiently completing the oncoming vehicle encounter.

[0025] In the above way, compared with the defect that the traditional step-by-step decision-making cannot achieve the global optimum, the technical solution of the present disclosure regards the ego vehicle and the oncoming vehicle as equal game players, can dynamically balance the behaviors of each vehicle, realizes more reasonable decision-making and planning in complex oncoming vehicle encounter scenarios, avoids the decision-making limitations brought by unilateral prediction, significantly improves the accuracy and reliability of oncoming vehicle trajectory prediction, enhances the system's ability to cope with dynamic environmental changes, and effectively guarantees the safety and efficiency of driving.

[0026] The following describes the present disclosure through specific embodiments in combination with specific application scenarios.

[0027] Please refer to Figure 1 , Figure 1 which is a flowchart of an oncoming vehicle trajectory prediction method shown in an exemplary embodiment. The method may perform the following steps: Step 102: In response to detecting an oncoming vehicle, traverse the behavior samples in the pre-constructed driving paradigm behavior space, where the behavior samples include the longitudinal acceleration and lateral offset parameters of the vehicle.

[0028] For example, the host vehicle (Vehicle A) is traveling at a speed of 60 km / h on a two-lane two-way road. At this time, an oncoming vehicle (Vehicle B) is detected in the oncoming lane, approaching at a speed of 70 km / h, and the distance between the two vehicles is 100 m. The system of Vehicle A triggers the oncoming vehicle trajectory prediction method and traverses the behavior samples from the pre-constructed driving paradigm behavior space. Suppose the behavior space contains the following behavior samples: Sample 1: Longitudinal acceleration = 0 m / s² (maintain a constant speed), lateral offset = 0.2 m (offset 0.2 m to the right side of the lane); Sample 2: Longitudinal acceleration = 0.1 m / s² (accelerate), lateral offset = -0.1 m (offset 0.1 m to the left side of the lane); Sample 3: Longitudinal acceleration = -0.3 m / s² (decelerate), lateral offset = -0.3 m (offset 0.3 m to the left side of the lane).

[0029] Among them, during the driving process of the host vehicle, when a vehicle is detected in the oncoming lane by the in-vehicle sensor system (such as cameras, radars, lidars, etc.), the oncoming vehicle trajectory prediction process is triggered. Before the prediction process starts, the system has constructed a driving paradigm behavior space through a large amount of real driving data. Specifically, the system has collected a large amount of driving data of vehicles in oncoming vehicle scenarios, and these data include the longitudinal acceleration of the vehicle (indicating the degree of acceleration or deceleration of the vehicle along the lane direction) and the lateral offset parameters (indicating the lateral position change of the vehicle relative to the center line of the lane).

[0030] In an illustrated embodiment, before traversing the behavior samples in the pre-constructed driving paradigm behavior space, the method further includes: obtaining a vehicle real driving data set, where each data in the vehicle real driving data set includes the longitudinal acceleration and lateral offset parameters of vehicle oncoming; screening the data in the vehicle real driving data set, only retaining the representative data of different driving modes, and constructing a driving paradigm behavior space according to the representative data of different driving modes.

[0031] For example, the system screens the driving data of a large number of vehicles in the oncoming vehicle scenario (the data includes the longitudinal acceleration and lateral offset parameters of the vehicle) by means of cluster analysis to obtain several cluster centers, and then selects the data near the cluster center points as representative item data. For example, 3 representative item data, namely longitudinal acceleration = 0 m / s² (maintaining a constant speed), lateral offset = 0.2 m (offset 0.2 m to the right side of the lane); longitudinal acceleration = 0.1 m / s² (accelerating), lateral offset = -0.1 m (offset 0.1 m to the left side of the lane); longitudinal acceleration = -0.3 m / s² (decelerating), lateral offset = -0.3 m (offset 0.3 m to the left side of the lane). Each type of representative item data may represent a driving style, such as aggressive, conservative, steady, etc. According to each representative item data, a driving paradigm behavior space is constructed.

[0032] Among them, the driving paradigm behavior space is constructed by obtaining a real driving data set and screening representative item data of different driving modes. These real driving data include the longitudinal acceleration and lateral offset parameters when the vehicles meet, and can truly reflect the behavior characteristics of the vehicles under different driving styles. By screening and retaining representative data, the constructed driving paradigm behavior space can not only cover various driving modes, but also maintain the simplicity of the data, improving the efficiency of subsequent traversal and deduction. Cluster analysis is to divide a large number of similar data points into groups to form different clusters. In this process, the system first defines the number of clusters to be generated, and then applies a clustering algorithm to try to minimize the distance between data points within the same cluster and maximize the distance between data points in different clusters. For each cluster, its geometric center, that is, the cluster center point, is calculated, which represents the average value of all data points within the cluster. Subsequently, the system selects the data points close to these cluster center points as representative items to ensure that the selected data can accurately reflect the main behavior characteristics of this driving mode.

[0033] It should be noted that in addition to screening the data in the real vehicle driving data set by means of cluster analysis, the data in the real vehicle driving data set can also be screened by means of manually formulating rules. For example, different threshold ranges of driving modes can be set according to traffic regulations, driving habits or human observations. The data items in the original data set that meet the above rules are classified into the corresponding driving style categories, and typical samples are selected from each category as representative item data. In addition, the data in the real vehicle driving data set can also be screened by means of deep learning. For example, the original driving data (including longitudinal acceleration and lateral offset) is constructed into a time series or state vector form and input into a neural network. The neural network uses a supervised deep learning model to classify the driving mode, and for each category, typical samples are selected as representative item data. The present disclosure does not limit the screening methods for the representative item data of different driving modes.

[0034] In related technologies, for the trajectory prediction of vehicles during oncoming vehicle encounters, it is usually difficult to balance sampling completeness and computational efficiency. Because fully enumerating all possible behavior combinations of vehicles will lead to excessive consumption of computing resources. If the computing power burden is reduced by decreasing the sampling quantity, key scenarios may be missed.

[0035] Through the technical means in this embodiment, various data in the vehicle's real driving dataset are screened to construct a behavior space with sparse completeness and conforming to driving paradigms, only retaining necessary representative data items, which helps to identify and distinguish different driving styles, and can effectively reduce redundant data and improve computational efficiency.

[0036] Step 104: After traversing to a behavior sample each time, based on the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively to obtain a trajectory pair of the host vehicle and the oncoming vehicle.

[0037] For example, after traversing to Sample 1, perform trajectory deduction on Vehicle A and Vehicle B according to Sample 1 combined with the current motion parameters of Vehicle A and Vehicle B. It can be known that next, Vehicle A will maintain a constant speed of 60 km / h and at the same time offset 0.2 m to the right side of the lane; Vehicle B will maintain a constant speed of 70 km / h and at the same time the lateral offset of Vehicle B is 0 (continue to drive at the original lateral position of the lane). According to the above trajectory deduction process, Trajectory Pair 1 of Vehicle A and Vehicle B is obtained. After traversing to Sample 2 and Sample 3, in a similar process as above, Trajectory Pair 2 and Trajectory Pair 3 of Vehicle A and Vehicle B are obtained.

[0038] Among them, when an oncoming vehicle is detected, the system starts to traverse the behavior samples in the driving paradigm behavior space. Each behavior sample contains specific longitudinal acceleration and lateral offset parameters, representing a possible driving behavior. The system will perform deduction on the motion trajectories of the host vehicle and the oncoming vehicle in the next period of time based on the current motion states (such as position, speed, etc.) of the host vehicle and the oncoming vehicle and the traversed behavior sample. The deduction method can be based on the target oncoming vehicle speed combined with the kinematic equation. For safe oncoming vehicle encounter, the target oncoming vehicle speed of the vehicle is usually a relatively low speed. If the current motion speed of the vehicle , the vehicle needs to perform a deceleration operation, select the longitudinal acceleration from the driving paradigm behavior space, and the deduction step size is , then the speed at the th time step, and the position at the th time step. When decelerating to the target oncoming vehicle speed When it enters the constant-speed passing stage. In rare cases, if the current moving speed of the vehicle , the vehicle needs to perform an acceleration operation, and select the longitudinal acceleration from the driving paradigm behavior space , and the derivation process is similar to the foregoing. If the current moving speed of the vehicle , it will perform a constant-speed movement until the passing ends.

[0039] After deriving the position at the th time step, based on the derived position, the lateral offset of the vehicle at the th time step can be derived in combination with the lateral offset parameter (including the lateral offset amount or the lateral offset angle). Among them, if the lateral offset parameter is the lateral offset amount, the lateral offset amount of the vehicle at the th time step can be directly obtained. If the lateral offset parameter is the lateral offset angle, the lateral offset amount of the th time step relative to the th time step , where is the longitudinal position deviation between two adjacent time steps, , is the longitudinal position of the vehicle at the th time step, is the longitudinal position of the vehicle at the th time step, is the th time step's lateral offset angle, which is selected from the driving paradigm behavior space.

[0040] It should be noted that in addition to using the above kinematic calculation formulas for trajectory derivation, a vehicle dynamics model can also be used. Combining the current state of the vehicle (speed, acceleration, steering angle, etc.), environmental constraints (road boundaries, obstacle positions), and the traversed behavior samples, the future trajectory of the vehicle is solved by numerical integration methods; or, a deep learning model is used to predict the future trajectory of the vehicle. For the method of separately performing trajectory derivation on the host vehicle and the oncoming vehicle according to the behavior samples and the current motion parameters of the host vehicle and the oncoming vehicle, the present disclosure does not limit this.

[0041] Step 106: Quantitatively evaluate a number of trajectory pairs obtained after the traversal ends to determine the target trajectory pair that fits the host vehicle.

[0042] For example, from the index dimensions such as safety, comfort, and passing efficiency, the system uses the evaluation functions of safety, comfort, and efficiency, and the weights of each function to perform weighted quantitative evaluation on the 3 groups of trajectory pairs obtained after the traversal ends. After evaluation, the trajectory pair of Sample 1 has the highest score and is determined as the target trajectory pair that fits Vehicle A.

[0043] Among them, by traversing all behavior samples in the driving paradigm behavior space, the system can obtain multiple possible trajectory pairs. Each trajectory pair includes the deduced trajectory of the host vehicle and the deduced trajectory of the oncoming vehicle. The quantitative evaluation is based on multiple indexes related to vehicle meeting, such as safety, comfort, efficiency, etc. By weighted calculation, the comprehensive score of each trajectory pair is obtained, so as to select the target trajectory pair most suitable for the host vehicle. The quantitative evaluation is based on multiple indexes related to vehicle meeting, such as safety, comfort, efficiency, etc. Each index has a corresponding evaluation function and weight. The evaluation function can be a cost function or a reward function. The reward function is usually the opposite value of the cost function. The cost function can include but is not limited to: the lateral safety cost function with the oncoming vehicle, the lateral offset cost function, the passability cost function, the distance cost function from obstacles, the acceleration comfort cost function, and the acceleration change rate comfort cost function, etc.

[0044] For example, the evaluation function of safety may be a function proportional to the minimum distance from the oncoming vehicle; the evaluation function of comfort may be a function inversely proportional to the acceleration change rate; the evaluation function of efficiency may be a function inversely proportional to the time taken to complete the vehicle meeting. The weight of each index reflects the relative importance of this index. In different driving scenarios, the weights of each index may be different. For example, on a narrow road, the weight of safety may be higher; on a wide road, the weight of efficiency may be higher.

[0045] It should be noted that the quantitative evaluation is not only limited to evaluation functions in different dimensions, but also a series of rules and thresholds can be manually formulated according to principles such as driving safety, comfort, and efficiency, or the behavior can be scored through reinforcement learning. Regarding the specific evaluation method and weight setting method, the present disclosure does not limit this.

[0046] Step 108: Control the host vehicle to drive according to the deduced trajectory of the host vehicle in the target trajectory pair.

[0047] For example, the system controls vehicle A to drive according to the deduced trajectory of the host vehicle in the target trajectory pair, that is, vehicle A maintains a constant speed of 60 km / h and simultaneously offsets 0.2 m to the right side of the lane, and then completes the vehicle meeting safely, comfortably and efficiently.

[0048] Among them, the target trajectory pair refers to the appropriate trajectory pair determined after quantitative evaluation among multiple possible trajectory pairs. This appropriate trajectory pair not only considers safety (such as avoiding collisions), but also takes into account comfort (such as smooth acceleration changes and lateral offsets) and efficiency (such as minimizing driving time and energy consumption). The deduced trajectory of the host vehicle is the predicted future movement path of the host vehicle based on the current driving situation and the selected behavior samples. Specifically, it includes motion parameters such as the speed, acceleration and lateral offset of the host vehicle in the future period of time.

[0049] In one of the illustrated embodiments, after each traversal to a behavior sample, according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, trajectory deduction is performed on the host vehicle and the oncoming vehicle respectively to obtain a trajectory pair of the host vehicle and the oncoming vehicle, including: after each traversal to a behavior sample, according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, trajectory deduction is performed on the host vehicle and the oncoming vehicle respectively for a trajectory deduction corresponding duration after the current moment; re-traverse the behavior samples in the driving paradigm behavior space, and when another behavior sample is traversed, repeat the above trajectory deduction process until the deduced trajectory pair is sufficient for the host vehicle and the oncoming vehicle to end the passing operation.

[0050] For example, vehicle A is traveling at a constant speed of 60 km / h, while vehicle B is traveling at a constant speed of 70 km / h, and the two vehicles are 30 m apart at this time. The system traverses to a behavior sample from a pre-constructed driving paradigm behavior space. This sample is longitudinal acceleration = 0 m / s² (maintaining a constant speed), and lateral offset = -0.3 m (offset 0.3 m to the left side of the lane). According to this sample, the system starts the first step of trajectory deduction. In the next deduction step of 0.1 s in the future, vehicle A maintains a speed of 60 km / h and offsets 0.3 m to the left side of the lane; at the same time, vehicle B also maintains a speed of 70 km / h and offsets 0.3 m to the left side of the lane. After the deduction step of 0.1 s ends, the distance between the two vehicles is shortened to approximately 26.39 m (considering the relative speed and time). Then, the system continues to traverse the next behavior sample from the pre-constructed driving paradigm behavior space and repeats the above trajectory deduction process until the distance between vehicle A and vehicle B is 0 m or less than 0 m, and it is considered that vehicle A and vehicle B have ended the passing operation.

[0051] Please refer to Figure 2 , Figure 2 which is a schematic diagram of vehicle trajectory deduction in a passing scenario shown in an exemplary embodiment. As Figure 2 shown, it shows a driving scenario of an autonomous vehicle in a complex traffic environment. There are two vehicles passing through a busy traffic area in the figure, and there are multiple stationary vehicles around. A passing operation is being carried out between the two vehicles to ensure safe passage. There are multiple arrows around the vehicles in the figure, representing the driving direction of the host vehicle (the vehicle on the left side in the figure) and the detection range of the on-vehicle sensors. The on-vehicle sensors can detect the distance between the two vehicles and the distance between the host vehicle and other obstacles, which are considered when performing trajectory planning and decision-making. A total of three trajectory deduction processes are shown in the figure, which are represented by three shaded blocks in the driving direction in front of each vehicle. After the three trajectory deduction processes end, the host vehicle and the oncoming vehicle have ended the passing process.

[0052] Among them, the duration corresponding to one deduction can be one deduction step length. When the rear ends of the ego vehicle and the oncoming vehicle are staggered after the passing of the oncoming vehicle, it can be considered that the passing of the ego vehicle and the oncoming vehicle is completed. The trajectory deduction adopts a step-by-step deduction method, and the trajectory within each deduction time step (such as 0.1 s) is deduced each time. Specifically, the system first calculates the positions, speeds, and accelerations of the ego vehicle and the oncoming vehicle in the next time step according to the current motion states of the ego vehicle and the oncoming vehicle, as well as the longitudinal acceleration and lateral offset parameters in the traversed behavior samples. Then, the system traverses the behavior samples again, selects the next behavior sample, and continues to deduce the trajectory in the next time step. This process continues until the deduced trajectory covers the entire passing process. This step-by-step deduction method can more accurately simulate the dynamic changes during the passing process and improve the accuracy of trajectory prediction.

[0053] In an illustrated embodiment, the method further includes: estimating the duration required for the ego vehicle and the oncoming vehicle to complete passing, so as to determine at least the number of deduction processes required to be performed preset times according to the duration corresponding to one deduction and the duration required for the passing to end; if the deduced trajectory pair has completed the deduction process for the preset number of times, it is determined that the deduced trajectory pair is sufficient for the ego vehicle and the oncoming vehicle to complete passing.

[0054] For example, to determine the duration for which the deduction needs to continue, the system estimates the duration required to complete passing based on factors such as the current speeds, positions of the ego vehicle and the oncoming vehicle, and the road width. For example, if the relative speed of the ego vehicle and the oncoming vehicle is 20 m / s and the current distance is 100 m, it takes approximately 5 s to complete passing. If the time step for one deduction is 0.1 s, then at least 50 deductions are required to cover the entire passing process. If the deduced trajectory pair has completed 50 deduction processes, then it can be considered that the trajectory pair can cover the entire passing process.

[0055] Among them, to ensure that the deduced trajectory pair can cover the entire passing process, this method estimates the duration required for the ego vehicle and the oncoming vehicle to complete passing, and calculates at least the number of deductions required according to the duration corresponding to one deduction. After the deduction process for the preset number of times is completed, it can be considered that the deduced trajectory pair is sufficient for the ego vehicle and the oncoming vehicle to complete passing.

[0056] Considering that the passing process is usually a deceleration process, when the distance between the two vehicles is fixed, the actual passing duration is longer than the normal driving duration, and there may be other unexpected situations or calculation errors. Therefore, the system generally increases a certain number of redundant times based on the calculated number of deductions. For example, the number of deductions is increased from 50 times to 100 times or more to ensure that the deduced trajectory pair can cover the entire passing process.

[0057] In one of the illustrated embodiments, after controlling the host vehicle to travel along the deduced trajectory of the host vehicle according to the target trajectory pair, the method further includes: if the passing process between the host vehicle and the oncoming vehicle has not been completed, predicting the ideal position of the oncoming vehicle according to the deduced trajectory of the oncoming vehicle in the target trajectory pair; detecting the deviation between the actual position and the ideal position of the oncoming vehicle, and if the deviation exceeds a preset deviation threshold, traversing the behavior samples in the pre-constructed driving paradigm behavior space again to determine the target trajectory pair suitable for the host vehicle again; if the deviation does not exceed the preset deviation threshold, continue to control the host vehicle to travel along the deduced trajectory of the host vehicle in the target trajectory pair.

[0058] For example, through comprehensive scoring, the system selects the trajectory pair with the highest score as the target trajectory pair and controls Vehicle A to travel along the deduced trajectory of the host vehicle in the target trajectory pair. During the control process, the system continuously monitors the deviation between the actual position of the oncoming vehicle and the predicted ideal position. If the deviation exceeds a preset threshold (such as 0.5 m), it means that the actual behavior of the oncoming vehicle does not match the prediction. At this time, the system traverses the behavior samples in the pre-constructed driving paradigm behavior space again and performs trajectory deduction according to the sample and the current motion parameters of the vehicle to adapt to the behavior changes of the oncoming vehicle.

[0059] Among them, during the process of controlling the host vehicle to travel along the deduced trajectory of the host vehicle in the target trajectory pair, the method also continuously monitors the deviation between the actual position of the oncoming vehicle and the predicted ideal position. If the deviation exceeds the preset deviation threshold, it means that there is a large difference between the actual driving trajectory of the oncoming vehicle and the predicted trajectory. At this time, the behavior samples will be traversed again and trajectory deduction will be performed to adapt to the behavior changes of the oncoming vehicle. If the deviation is within an acceptable range, continue to travel along the original trajectory.

[0060] In the related art, setting the trajectory prediction frequency for autonomous vehicles usually faces a dilemma. High-frequency decision-making will increase the computational resource load and cause frequent switching of vehicle behavior, reducing the actual position of ride comfort and safety; however, blindly reducing the decision-making frequency will cause environmental response delay and decision lag. In this embodiment, after the deviation between the predicted ideal position of the oncoming vehicle and the actual position of the oncoming vehicle exceeds the preset threshold, trajectory prediction is performed again, so as to maintain an appropriate decision-making frequency and maintain the safety and comfort of driving.

[0061] In an illustrated embodiment, controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair includes: controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair and accumulating the number of decision steps once; after controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: if the passing process between the host vehicle and the oncoming vehicle has not been completed, counting the accumulated number of decision steps; if the accumulated number of decision steps reaches a preset step threshold, traversing the behavior samples in the pre-constructed driving paradigm behavior space again to re-determine the target trajectory pair suitable for the host vehicle; if the accumulated number of decision steps does not reach the preset step threshold, continuing to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair and accumulating the number of decision steps once.

[0062] For example, please refer to Figure 3 , Figure 3 which is a flowchart of another passing trajectory prediction method shown in an exemplary embodiment. As Figure 3 shown, this flowchart details how the host vehicle predicts and executes the optimal trajectory through a series of steps after detecting the oncoming vehicle to safely and efficiently complete the passing process. The process starts with detecting the oncoming vehicle, and then initializes the number of decision steps j to 0. Next, the system performs trajectory prediction, including behavior sampling, trajectory deduction, trajectory pruning, and trajectory selection. In the trajectory prediction stage, the system deduces trajectories based on the behavior samples traversed in the pre-constructed driving paradigm behavior space and the current motion parameters of the host vehicle and the oncoming vehicle, obtaining multiple possible trajectory pairs. Trajectory pruning is performed on all possible trajectory pairs to remove the trajectory pairs with collision risks, and then the remaining trajectory pairs are quantitatively evaluated to obtain the optimal trajectory pair suitable for the host vehicle. The optimal trajectory pair is executed for the host vehicle, and the initialized number of decision steps j is incremented by 1. During the travel of the host vehicle, it is determined whether the passing process with the oncoming vehicle is completed. If the passing is not completed, the system detects whether the initialized number of decision steps j reaches the preset number of steps N, and collects the actual position of the oncoming vehicle, compares it with the predicted position of the oncoming vehicle determined by the deduced trajectory of the oncoming vehicle, and determines whether the position deviation exceeds the threshold. If either of these two conditions is met, it means that the currently deduced trajectory of the host vehicle is no longer suitable for continued use, and the system will re-perform trajectory prediction to adapt to the possibly changing traffic conditions. If neither of these two conditions is met, the system will continue to execute the currently deduced trajectory of the host vehicle and increment the number of decision steps again. The above process will continue until the passing is completed, at which time the system will restore the host vehicle to the cruise mode and continue to detect whether there is an oncoming vehicle.

[0063] Among them, the concept of decision-making steps is introduced in this embodiment. Each time the system controls the ego vehicle to travel along the deduced trajectory for a time step, the decision-making steps are accumulated once. When the accumulated decision-making steps reach the preset step threshold, the trajectory deduction is restarted to ensure the timeliness and accuracy of the control strategy. This dynamic adjustment mechanism enables the system to flexibly respond to complex traffic environments and improve the safety and efficiency of the oncoming vehicle meeting process.

[0064] In one illustrated embodiment, the method further includes: before quantifying and evaluating a number of trajectory pairs obtained after the traversal to determine a target trajectory pair suitable for the ego vehicle, the method further includes: performing a collision risk assessment on a number of trajectory pairs obtained after the traversal according to a preset risk assessment rule; and retaining only a number of trajectory pairs with qualified collision risk assessment results according to the collision risk assessment result.

[0065] For example, the system obtains three groups of trajectory pairs after the traversal, namely trajectory pair A, trajectory pair B, and trajectory pair C. Each group of trajectory pairs includes different driving paths that the ego vehicle and the oncoming vehicle may take during the oncoming vehicle meeting process. Before performing the quantitative evaluation, the system will first perform a collision risk assessment on these trajectory pairs to screen out the trajectory pairs with potential safety hazards. The collision risk assessment can be carried out based on the risk assessment rule of the minimum safety distance: The system sets the minimum safety distance between vehicles and between a vehicle and the road boundary to 0.25 m. If during the oncoming vehicle meeting process, when the ego vehicle and the oncoming vehicle travel according to trajectory pair A, the distance between the ego vehicle and the oncoming vehicle, or the distance between any vehicle and the road boundary is less than 0.25 m, and when the ego vehicle and the oncoming vehicle travel according to trajectory pair B or trajectory pair C, the distance between the ego vehicle and the oncoming vehicle, or the distance between any vehicle and the road boundary is not less than 0.25 m, then trajectory pair A is determined to be unsafe, and only trajectory pair B and trajectory pair C are retained for the next quantitative evaluation.

[0066] Among them, before performing the quantitative evaluation, the system will first perform a collision risk assessment on the trajectory pairs to screen out the trajectory pairs with potential safety hazards, ensuring that the finally selected trajectory pairs can ensure driving safety. The collision risk assessment is carried out based on the preset risk assessment rules. These rules may include: the minimum safety distance between vehicles, the minimum safety distance between a vehicle and the road boundary, the maximum allowable acceleration and deceleration of the vehicle, etc. The system will check whether each trajectory pair violates these safety rules during the entire oncoming vehicle meeting process. If it violates, the trajectory pair is marked as unsafe and will not be further considered.

[0067] In addition, this embodiment also introduces a learning mechanism. The system will record the data and control effects during each oncoming vehicle meeting process and continuously optimize the driving paradigm behavior space and the evaluation method model with these data, enabling the system to continuously adapt to different driving environments and driving styles.

[0068] Through the above steps, the oncoming vehicle trajectory prediction method provided in this embodiment can accurately predict the trajectories of the host vehicle and the oncoming vehicle during the oncoming process, and control the host vehicle to complete the oncoming process safely, comfortably and efficiently according to the prediction results.

[0069] Corresponding to the embodiment of the above oncoming vehicle trajectory prediction method, the present disclosure also provides an embodiment of an oncoming vehicle trajectory prediction device.

[0070] Please refer to Figure 4 , Figure 4 which is a hardware structure diagram of an electronic device shown in an exemplary embodiment. At the hardware level, the device includes a processor 402, an internal bus 404, a network interface 406, a memory 408, and a non-volatile memory 410. Of course, other required hardware may also be included. One or more embodiments of the present disclosure can be implemented in a software manner. For example, the processor 402 reads the corresponding computer program from the non-volatile memory 410 into the memory 408 and then runs it. Of course, in addition to the software implementation manner, one or more embodiments of the present disclosure do not exclude other implementation manners, such as a logic device or a combination of software and hardware. That is to say, the execution subject of the following processing flow is not limited to each logic unit, and can also be hardware or a logic device.

[0071] Please refer to Figure 5 , Figure 5 which is a block diagram of an oncoming vehicle trajectory prediction device shown in an exemplary embodiment. The oncoming vehicle trajectory prediction device 500 can be applied to an electronic device as shown in Figure 4 to implement the technical solution of the present disclosure. The device includes: A traversal unit 502, configured to traverse behavior samples in a pre-constructed driving paradigm behavior space in response to detecting an oncoming vehicle, where the behavior samples include longitudinal acceleration and lateral offset parameters of the vehicle; A deduction unit 504, configured to, after traversing to a behavior sample each time, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, and obtain a trajectory pair of the host vehicle and the oncoming vehicle; A first evaluation unit 506, configured to quantitatively evaluate a number of trajectory pairs obtained after the traversal ends to determine a target trajectory pair that fits the host vehicle; A first control unit 508, configured to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

[0072] In some embodiments, before traversing the behavior samples in the pre-constructed driving paradigm behavior space, the device further includes: An acquisition unit 510, configured to acquire a vehicle real driving data set, where each piece of data in the vehicle real driving data set includes the longitudinal acceleration and lateral offset parameters of vehicle meeting; A screening unit 512, configured to screen each piece of data in the vehicle real driving data set, only retain the representative item data of different driving modes, and construct a driving paradigm behavior space according to the representative item data of different driving modes.

[0073] In some embodiments, after traversing to a behavior sample each time, according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, trajectory deduction is respectively performed on the host vehicle and the oncoming vehicle to obtain a trajectory pair of the host vehicle and the oncoming vehicle, including: After traversing to a behavior sample each time, according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, trajectory deduction is respectively performed on the host vehicle and the oncoming vehicle for a corresponding duration after the current moment; Re-traverse the behavior samples in the driving paradigm behavior space. When traversing to another behavior sample, repeat the above trajectory deduction process until the deduced trajectory pair is sufficient for the host vehicle and the oncoming vehicle to end the meeting.

[0074] In some embodiments, the apparatus further includes: An estimation unit 514, configured to estimate the duration required for the host vehicle and the oncoming vehicle to end the meeting, so as to determine at least the number of preset deduction processes according to the corresponding duration of one deduction and the duration required for ending the meeting; A first determination unit 516, configured to determine that the deduced trajectory pair is sufficient for the host vehicle and the oncoming vehicle to end the meeting if the deduced trajectory pair has completed the preset number of deduction processes.

[0075] In some embodiments, after controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the apparatus further includes: A prediction unit 518, configured to predict the ideal position of the oncoming vehicle according to the deduced trajectory of the oncoming vehicle in the target trajectory pair if the meeting process between the host vehicle and the oncoming vehicle has not been completed; A second determination unit 520, configured to detect the deviation between the actual position and the ideal position of the oncoming vehicle. If the deviation exceeds a preset deviation threshold, re-traverse the behavior samples in the pre-constructed driving paradigm behavior space to re-determine the target trajectory pair suitable for the host vehicle; A second control unit 522, configured to continue to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair if the deviation does not exceed the preset deviation threshold.

[0076] In some embodiments, controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair includes: controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, and accumulating one decision step count; After controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: If the passing process between the host vehicle and the oncoming vehicle has not been completed, then count the accumulated decision step count; If the accumulated decision step count reaches a preset step count threshold, then traverse the behavior samples in the pre-constructed driving paradigm behavior space again to determine the target trajectory pair suitable for the host vehicle again; If the accumulated decision step count does not reach the preset step count threshold, then continue to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, and accumulate one decision step count.

[0077] In some embodiments, before quantifying and evaluating the several trajectory pairs obtained after the traversal to determine the target trajectory pair suitable for the host vehicle, the device further includes: A second evaluation unit 524, configured to perform a collision risk assessment on the several trajectory pairs obtained after the traversal according to a preset risk assessment rule; A retention unit 526, configured to retain only the several trajectory pairs with qualified collision risk assessment results according to the collision risk assessment result.

[0078] The implementation processes of the functions and roles of each unit in the above device are specifically described in detail in the implementation processes of the corresponding steps in the above method, and will not be elaborated here.

[0079] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only illustrative. The units described as separate components may or may not be physically separated, and the components shown 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 modules can be selected according to actual needs to achieve the purpose of the present disclosure solution. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0080] The systems, devices, modules or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, and the specific form of the computer can be a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email transceiver device, a game console, a tablet computer, a wearable device, or a combination of any several of these devices.

[0081] In a typical configuration, a computer includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0082] The memory may include non-permanent memory in the computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0083] Computer-readable media includes permanent and non-permanent, removable and non-removable media that can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transmission media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0084] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data that have been authorized by the user or fully authorized by all parties. And the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0085] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising said element.

[0086] The specific embodiments of the present disclosure have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require the particular order or sequential order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0087] The terms used in one or more embodiments of the present disclosure are for the purpose of describing specific embodiments only and are not intended to limit one or more embodiments of the present disclosure. The singular forms "a", "the" and "said" used in one or more embodiments of the present disclosure and the appended claims are also intended to include the plural forms unless the context clearly dictates otherwise. It should also be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.

[0088] It should be understood that although the terms first, second, third, etc. may be used in one or more embodiments of the present disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of one or more embodiments of the present disclosure, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "when" or "while" or "in response to determining".

[0089] The above description is only the preferred embodiment of one or more embodiments of the present disclosure and is not intended to limit one or more embodiments of the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of one or more embodiments of the present disclosure shall be included within the scope of protection of one or more embodiments of the present disclosure.

Claims

1. A method for predicting the passing trajectory, characterized in that The method includes: In response to detecting an oncoming vehicle, traversing behavior samples in a pre-constructed driving paradigm behavior space, where the behavior samples include the longitudinal acceleration and lateral offset parameters of the vehicle; After traversing to a behavior sample each time, performing trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, to obtain a trajectory pair of the host vehicle and the oncoming vehicle; Quantitatively evaluating a number of trajectory pairs obtained after the traversal ends, to determine a target trajectory pair suitable for the host vehicle; Controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

2. The method according to claim 1, wherein Before traversing the behavior samples in the pre-constructed driving paradigm behavior space, the method further includes: Obtaining a vehicle real driving data set, where each data in the vehicle real driving data set includes the longitudinal acceleration and lateral offset parameters of vehicle meeting; Screening the data in the vehicle real driving data set, only retaining the representative item data of different driving modes, and constructing a driving paradigm behavior space according to the representative item data of different driving modes.

3. The method according to claim 1, wherein The step of, after traversing to a behavior sample each time, performing trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, to obtain a trajectory pair of the host vehicle and the oncoming vehicle, includes: After traversing to a behavior sample each time, performing trajectory deduction for a corresponding duration of one deduction after the current moment on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle; Re-traversing the behavior samples in the driving paradigm behavior space, and when traversing to another behavior sample, repeating the above trajectory deduction process until the deduced trajectory pairs are sufficient for the host vehicle and the oncoming vehicle to end the meeting.

4. The method according to claim 3, characterized in that The method further includes: Estimating the duration required for the host vehicle and the oncoming vehicle to end the meeting, to determine at least a preset number of deduction processes according to the corresponding duration of one deduction and the duration required for ending the meeting; If the deduced trajectory pairs have completed the preset number of deduction processes, it is determined that the deduced trajectory pairs are sufficient for the host vehicle and the oncoming vehicle to end the meeting.

5. The method according to claim 1, wherein After controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: If the meeting process between the host vehicle and the oncoming vehicle has not been completed, predicting the ideal position of the oncoming vehicle according to the deduced trajectory of the oncoming vehicle in the target trajectory pair; Detecting the deviation between the actual position and the ideal position of the oncoming vehicle, and if the deviation exceeds a preset deviation threshold, re-traversing the behavior samples in the pre-constructed driving paradigm behavior space to determine the target trajectory pair suitable for the host vehicle again; If the deviation does not exceed the preset deviation threshold, continue to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

6. The method according to claim 1 or 5, characterized in that, Controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair includes: controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, and accumulating one decision step count; After controlling the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, the method further includes: If the passing process between the host vehicle and the oncoming vehicle has not been completed, then count the accumulated decision step count; If the accumulated decision step count reaches a preset step count threshold, then traverse the behavior samples in the pre-constructed driving paradigm behavior space again to re-determine the target trajectory pair adapted to the host vehicle; If the accumulated decision step count does not reach the preset step count threshold, then continue to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair, and accumulate one decision step count.

7. The method according to claim 1, wherein Before quantifying and evaluating the several trajectory pairs obtained after the traversal ends to determine the target trajectory pair adapted to the host vehicle among them, the method further includes: Performing a collision risk assessment on the several trajectory pairs obtained after the traversal ends according to a preset risk assessment rule; According to the collision risk assessment result, only retain the several trajectory pairs with qualified collision risk assessment results.

8. A passing trajectory prediction device, characterized in that, The device includes: A traversal unit, configured to traverse the behavior samples in the pre-constructed driving paradigm behavior space in response to detecting an oncoming vehicle, where the behavior samples include the longitudinal acceleration and lateral offset parameters of the vehicle; A deduction unit, configured to, after traversing to a behavior sample each time, perform trajectory deduction on the host vehicle and the oncoming vehicle respectively according to the behavior sample and the current motion parameters of the host vehicle and the oncoming vehicle, to obtain the trajectory pair of the host vehicle and the oncoming vehicle; An evaluation unit, configured to perform a quantitative evaluation on the several trajectory pairs obtained after the traversal ends to determine the target trajectory pair adapted to the host vehicle among them; A control unit, configured to control the host vehicle to travel according to the deduced trajectory of the host vehicle in the target trajectory pair.

9. A vehicle, characterized in that, Includes: A processor; A memory for storing instructions executable by the processor; Wherein, the processor realizes the method according to any one of claims 1 to 7 by running the executable instructions.

10. A computer program product, characterized in that, Includes a computer program / instructions, and when the computer program / instructions are executed by the processor, the steps of the method according to any one of claims 1 to 7 are realized.

Images