Automatic driving decision-making method and device, electronic equipment and storage medium

Abstract

The invention provides an automatic driving decision-making method and device, electronic equipment and a storage medium, and relates to the technical field of automatic driving. The method comprises the steps that first motion information of a target obstacle vehicle is acquired, and a driving route conflict exists between the target obstacle vehicle and a vehicle; according to the first motion information, a robbing strategy is determined, and the robbing strategy comprises at least one of the following steps that the speed of the vehicle is increased, and the transverse vehicle distance between the vehicle and the target obstacle vehicle is shortened; acquiring second motion information of the target obstacle vehicle; on the basis of the second motion information, predicting the driving intention of the target obstacle vehicle after the robbing strategy is executed; and under the condition that the driving intention of the target obstacle vehicle is yielding, executing a robbing strategy. According to the embodiment of the invention, the driving game with the target obstacle vehicle can be carried out on the premise of the robbing behavior in the presence of the obstacle vehicle, so that the passing efficiency of the automatic driving vehicle is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of autonomous driving technology, and in particular to an autonomous driving decision-making method, device, electronic device and storage medium. Background Art

[0002] In autonomous driving technology, in order to ensure driving safety, autonomous vehicles are usually more conservative in their choice of driving strategies. When an obstacle vehicle is detected near the autonomous vehicle, the autonomous vehicle will often give way to the obstacle vehicle excessively for safety reasons, resulting in low traffic efficiency of the autonomous vehicle and possible traffic jams. Summary of the invention

[0003] In order to overcome the problems existing in the related art, the present disclosure provides an automatic driving decision-making method, device, electronic device and storage medium.

[0004] According to a first aspect of an embodiment of the present disclosure, there is provided an automatic driving decision-making method, comprising:

[0005] Acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle;

[0006] Determining a rush-to-passenger strategy according to the first movement information;

[0007] Acquiring second motion information of the target obstacle vehicle;

[0008] Based on the second motion information, predicting the driving intention of the target obstacle vehicle after the overtaking strategy is executed;

[0009] When the target obstacle vehicle has the driving intention of yielding, the overtaking strategy is executed.

[0010] In some embodiments, the method further comprises:

[0011] Predicting the driving routes of all vehicles in the driving environment of the self-vehicle;

[0012] If there are multiple obstacle vehicles among all the vehicles whose driving routes conflict with the own vehicle, predict the moments when the multiple obstacle vehicles conflict with the own vehicle respectively;

[0013] The obstacle vehicle which collides with the own vehicle earliest among the plurality of obstacle vehicles is used as the target obstacle vehicle.

[0014] In some embodiments, according to the first motion information, the rushing strategy is determined as: controlling the speed of the ego vehicle to a first speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a first lateral distance;

[0015] The first speed is positively correlated with the speed of the target obstacle vehicle, and the first lateral distance is negatively correlated with the longitudinal distance between the ego vehicle and the target obstacle vehicle.

[0016] In some embodiments, the predicting, based on the second motion information, the driving intention of the target obstacle vehicle after the execution of the rushing strategy includes:

[0017] The second motion information is combined with the overtaking strategy and input into a pre-trained intention prediction model to obtain the driving intention of the target obstacle vehicle, wherein the intention prediction model is used to characterize the mapping relationship between the second motion information and the driving intention under different overtaking strategies; or

[0018] In a pre-constructed driving intention table, the driving intention corresponding to the second motion information and the overtaking strategy is queried, wherein the driving intention table is used to record the mapping relationship between the second motion information and the driving intention under different overtaking strategies.

[0019] In some embodiments, the method further comprises:

[0020] In the case where the target obstacle vehicle has a driving intention of not yielding, determining a yielding strategy;

[0021] The yield strategy is executed.

[0022] In some embodiments, determining the yield strategy includes:

[0023] According to the first motion information, the yielding strategy is determined as: controlling the speed of the ego vehicle to a second speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a second lateral distance;

[0024] The second speed is less than the speed of the target obstacle vehicle, and the second lateral distance is greater than or equal to the current lateral distance between the ego vehicle and the target obstacle vehicle.

[0025] In some embodiments, after executing the rush-to-go strategy, the method further includes:

[0026] If there is still a driving route conflict between the target obstacle vehicle and the self-vehicle, the overtaking strategy is determined again, the driving intention of the target obstacle vehicle after the overtaking strategy is executed is predicted, and the overtaking strategy is executed when the driving intention is to give way.

[0027] According to a second aspect of an embodiment of the present disclosure, there is provided an automatic driving decision-making device, comprising:

[0028] A first acquisition module is used to acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle;

[0029] A determination module, configured to determine a preemptive strategy according to the first motion information, wherein the preemptive strategy includes at least one of the following: increasing the speed of the ego vehicle, shortening the lateral distance between the ego vehicle and the target obstacle vehicle;

[0030] A second acquisition module, used to acquire second motion information of the target obstacle vehicle;

[0031] A prediction module, configured to predict the driving intention of the target obstacle vehicle after the execution of the rushing strategy based on the second motion information;

[0032] The execution module is used to execute the overtaking strategy when the target obstacle vehicle has the driving intention of giving way.

[0033] According to a third aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0034] Processor; and

[0035] A memory, configured to store executable instructions of the processor;

[0036] The processor is configured to execute the method described in the first aspect by executing the executable instructions.

[0037] According to a fourth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the method described in the first aspect is implemented.

[0038] The solution provided by the embodiment of the present disclosure can obtain the first motion information of the target obstacle vehicle, and determine the overtaking strategy based on the first motion information. Among them, the overtaking strategy includes at least one of the following: increasing the speed of the own vehicle, shortening the lateral distance between the own vehicle and the target obstacle vehicle. Next, the second motion information of the target obstacle vehicle can be obtained, and based on the second motion information, the driving intention of the target obstacle vehicle after the execution of the overtaking strategy can be predicted. In the case that the driving intention of the target obstacle vehicle is to give way, the overtaking strategy is executed. The embodiment of the present disclosure can, in the presence of an obstacle vehicle, conduct a driving game with the target obstacle vehicle on the premise of overtaking, thereby improving the traffic efficiency of the autonomous driving vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 A flow chart of an autonomous driving decision-making method in an embodiment of the present disclosure is shown.

[0040] Figure 2A schematic diagram of a driving scenario of a vehicle changing lanes in an embodiment of the present disclosure is shown.

[0041] Figure 3 A schematic diagram of a driving scenario of merging from a ramp into a main road in an embodiment of the present disclosure is shown.

[0042] Figure 4 A schematic diagram of the structure of an automatic driving decision-making device according to an embodiment of the present disclosure is shown.

[0043] Figure 5 A schematic structural diagram of an electronic device in an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0044] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0045] In the current mainstream autonomous driving architecture, the obstacle prediction module and the autonomous driving decision module are decoupled. The obstacle prediction module is responsible for predicting the behavior of obstacles and generating future driving trajectories, while the autonomous driving decision module is responsible for the behavioral decision of the autonomous driving vehicle. The important input for its behavioral decision is the obstacle prediction information output by the prediction module. Therefore, in scenarios where there is a strong game relationship between obstacles and autonomous driving vehicles, the driving behavior of autonomous driving vehicles will be too conservative and excessively give way to obstacles, causing the autonomous driving vehicle to stagnate and possibly block traffic.

[0046] Among them, the above-mentioned scenarios with a strong game relationship can be scenarios where there is a driving route conflict between the target obstacle vehicle and the autonomous driving vehicle, for example, driving scenarios such as simultaneous release of straight driving and left turns at an intersection, merging from a ramp into the main road, and vehicle lane changes.

[0047] In view of this, an automatic driving decision-making method is provided in an embodiment of the present disclosure. The method can be executed by an automatic driving vehicle, and can play a driving game with a target obstacle vehicle on the premise of the self-vehicle grabbing behavior, thereby improving the traffic efficiency of the automatic driving vehicle.

[0048] Figure 1 A flowchart of an automatic driving decision-making method in an embodiment of the present disclosure is shown as follows: Figure 1 As shown, the autonomous driving decision-making method provided in the embodiment of the present disclosure includes the following steps S101 to S105.

[0049] S101, obtaining first motion information of a target obstacle vehicle.

[0050] It should be noted that the target obstacle vehicle is a vehicle that has a driving route conflict with the ego vehicle. The ego vehicle is the autonomous driving vehicle itself, and the ego vehicle can decide to execute a rushing strategy or a yielding strategy based on the movement information of the target obstacle vehicle to resolve the driving route conflict between the target obstacle vehicle and the ego vehicle.

[0051] The driving route conflict can be understood as the driving route of the target obstacle vehicle and the driving route of the ego vehicle will intersect at a certain moment, that is, the target obstacle vehicle and the ego vehicle will collide at that moment.

[0052] In some embodiments, the first motion information of the target obstacle vehicle may include at least one of the following:

[0053] The location of the target obstacle vehicle;

[0054] The speed of the target obstacle vehicle;

[0055] The acceleration of the target obstacle vehicle; and

[0056] The angle between the target obstacle vehicle and the lane it is in.

[0057] Exemplarily, the first motion information of the target obstacle vehicle can be collected by a sensor configured on the vehicle, such as a binocular camera, a speed measuring radar, a laser radar, etc., which will not be described in detail in the embodiments of the present disclosure.

[0058] S102: Determine a rush-to-go strategy based on the first movement information.

[0059] Among them, the rushing strategy includes at least one of the following: increasing the speed of the vehicle, shortening the lateral distance between the vehicle and the target obstacle vehicle. For example, in the scenario where the vehicle expects to change lanes to a designated lane, if the target obstacle vehicle is in the designated lane and is behind the vehicle, the vehicle can try to shorten the lateral distance between the vehicle and the target obstacle vehicle to slow down the target obstacle vehicle and allow the vehicle to complete the lane change. For another example, when the target obstacle vehicle merges into the main road where the vehicle is located from the ramp, the vehicle can try to increase the speed of the vehicle to prompt the target obstacle vehicle to give way to the vehicle so that the vehicle can pass first.

[0060] In some embodiments, the rushing strategy may be determined according to the first motion information as: controlling the speed of the ego vehicle to a first speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a first lateral distance.

[0061] Among them, the first speed is positively correlated with the speed of the target obstacle vehicle, and the first lateral distance is negatively correlated with the longitudinal distance between the ego vehicle and the target obstacle vehicle.

[0062] That is to say, from the perspective of speed, in order for the ego vehicle to successfully overtake, the faster the target obstacle vehicle is, the faster the ego vehicle should be. From the perspective of lateral distance, the longer the longitudinal distance between the ego vehicle and the target obstacle vehicle is, the smaller the risk of shortening the lateral distance is. Therefore, the lateral distance can be made negatively correlated with the longitudinal distance to achieve the purpose of forcing the target obstacle vehicle to slow down, thereby completing the overtaking.

[0063] In some embodiments, the first motion information includes the position and speed of the target obstacle vehicle. In this case, the way to determine the rushing strategy can be:

[0064] If the longitudinal position of the target obstacle vehicle is in front of the ego vehicle, the rushing strategy is to increase the speed of the ego vehicle.

[0065] This determination method corresponds to the scenario where the target obstacle vehicle merges into the lane where the ego vehicle is located. It is understandable that when the target obstacle vehicle merges in front of the ego vehicle, there may be a driving route conflict between the ego vehicle and the normally traveling ego vehicle. At this time, the ego vehicle can increase its own speed and rush through to improve its own traffic efficiency.

[0066] If the longitudinal position of the target obstacle vehicle is behind the ego vehicle and the speed of the ego vehicle is greater than that of the target obstacle vehicle, the overtaking strategy is to shorten the lateral distance between the ego vehicle and the target obstacle vehicle.

[0067] This determination method corresponds to the scenario where the ego vehicle merges into the lane where the target obstacle vehicle is located. It is understandable that when the ego vehicle merges in front of the target obstacle vehicle, there may be a driving route conflict with the normally traveling target obstacle vehicle. At this time, if the speed of the ego vehicle is already greater than the speed of the target obstacle vehicle, the lateral distance between the ego vehicle and the target obstacle vehicle can be shortened to cause the target obstacle vehicle behind to slow down, so that the ego vehicle has priority.

[0068] If the longitudinal position of the target obstacle vehicle is behind the ego vehicle and the speed of the ego vehicle is less than or equal to the speed of the target obstacle vehicle, the rushing strategy is to increase the speed of the ego vehicle and shorten the lateral distance between the ego vehicle and the target obstacle vehicle.

[0069] This determination method also corresponds to the scenario where the ego vehicle merges into the lane where the target obstacle vehicle is located. Similarly, if the speed of the ego vehicle is less than or equal to the speed of the target obstacle vehicle, the ego vehicle will not be able to complete the lane change and overtake, and the driver of the target obstacle vehicle will find it difficult to perceive the ego vehicle's intention to change lanes and overtake. At this time, the lateral distance between the ego vehicle and the target obstacle vehicle can be shortened while increasing the speed of the ego vehicle so that the driver of the target obstacle vehicle behind can perceive the ego vehicle's intention, prompting the target obstacle vehicle to slow down and giving the ego vehicle priority.

[0070] In some embodiments, the determination of the overtaking strategy can also be achieved through a neural network model. For example, the first motion information can be input into a pre-trained strategy determination model to predict the overtaking strategy that the vehicle can effectively execute in the current state, which is not described in detail in the present embodiment.

[0071] S103, obtaining second motion information of the target obstacle vehicle.

[0072] Exemplarily, similar to the first motion information, the second motion information may also include at least one of the following:

[0073] The location of the target obstacle vehicle;

[0074] The speed of the target obstacle vehicle;

[0075] The acceleration of the target obstacle vehicle; and

[0076] The angle between the target obstacle vehicle and the lane it is in.

[0077] The method for collecting the second motion information is the same as that for the first motion information, and is not elaborated in detail in the embodiment of the present disclosure.

[0078] It is worth noting that the parameter types included in the second motion information may be partially the same, completely the same, or completely different from those in the first motion information. For example, the first motion information may be the position and speed of the target obstacle vehicle, while the second motion information may be the speed and acceleration of the target obstacle vehicle.

[0079] It is understandable that the first motion information is motion information used to determine the rushing strategy, while the second motion information is motion information used to predict the driving intention of the target obstacle vehicle. Therefore, the parameter type selection of the first motion information and the second motion information can be determined in combination with their actual use. For example, in the embodiment given in S102, the rushing strategy can be determined based on the position and speed of the target obstacle vehicle. In this case, the first motion information can select the position and speed of the target obstacle vehicle.

[0080] S104: predicting the driving intention of the target obstacle vehicle after the overtaking strategy is executed based on the second motion information.

[0081] In some embodiments, a functional relationship between the second motion information and the yield probability of the target obstacle vehicle after the execution of the rush strategy can be established. The functional relationship can be linear, nonlinear, or a complex functional relationship based on a neural network model. The functional relationship can predict the yield probability of the target obstacle vehicle after the execution of the rush strategy based on any one type of information or a combination of multiple types of information in the second motion information. For example, the yield probability can be predicted based on acceleration, the yield probability can be predicted based on speed and position, or the yield probability can be predicted based on speed and lane angle, and the disclosed embodiments do not limit this.

[0082] In some embodiments, the functional relationship can be represented as a pre-trained intention prediction model. The second motion information is combined with the preemptive strategy and input into the pre-trained intention prediction model to obtain the driving intention of the target obstacle vehicle. The intention prediction model is used to characterize the mapping relationship between the second motion information and the driving intention under different preemptive strategies.

[0083] Exemplarily, the intention prediction model can predict the probability of the target obstacle vehicle performing various response actions (for example, overtaking, giving way, maintaining the original state, etc.) after the ego vehicle executes the overtaking strategy based on the input second motion information and the overtaking strategy, and take the response action with the highest probability as the driving intention of the target obstacle vehicle.

[0084] Exemplarily, intention prediction can also be simplified as a binary classification problem of a neural network model, that is, the driving intention of the target obstacle vehicle is divided into "rushing" and "non-rushing", thereby completing the intention prediction of the target obstacle vehicle.

[0085] In some embodiments, different intention prediction models can also be trained for different rushing strategies. For example, the first intention prediction model is trained to predict the driving intention of the target obstacle vehicle after the rushing strategy of "increasing the speed of the own vehicle" is executed. The second intention prediction model is trained to predict the driving intention of the target obstacle vehicle after the rushing strategy of "shortening the lateral distance between the own vehicle and the target obstacle vehicle" is executed. The third intention prediction model is trained to predict the driving intention of the target obstacle vehicle when the above two rushing strategies are executed at the same time. This method can reduce the output parameters of a single model, thereby reducing the complexity of the model and improving the prediction accuracy.

[0086] Exemplarily, the intention prediction model may be trained by supervised learning, reinforcement learning, or other machine learning methods, so that the trained intention prediction model exhibits good performance and robustness in actual scenarios.

[0087] In some embodiments, the driving intention corresponding to the second motion information and the rushing strategy can be queried in a pre-constructed driving intention table. The driving intention table is used to record the mapping relationship between the second motion information and the driving intention under different rushing strategies. The data in the driving intention table can be obtained by designing experimental tests, or by collecting real traffic sample data statistics.

[0088] For example, a driving intention table including a mapping relationship between acceleration (or other second motion information) and driving intention can be constructed. Please refer to Table 1, which exemplarily shows a driving intention table, which can be divided into three columns, namely "self-vehicle overtaking strategy", "obstacle vehicle acceleration" and "obstacle vehicle driving intention". In Table 1, according to the overtaking strategy determined in S102, combined with the obstacle vehicle acceleration (second motion information), the driving intention of the obstacle vehicle can be matched.

[0089] Table 1 Driving intention table

[0090]

[0091] In some embodiments, the driving intention of the target obstacle vehicle predicted by any of the above methods can be divided into two situations: "yield" and "non-yield".

[0092] Among them, "giving way" means that after the execution of the rushing strategy, there is no conflict between the driving route of the target obstacle vehicle and the driving route of the own vehicle, or at least there is no conflict within a certain period of time (for example, 5 seconds).

[0093] “Non-yielding” means that after the execution of the overtaking strategy, the driving route of the target obstacle vehicle conflicts with the driving route of the ego vehicle within a certain period of time (for example, 5 seconds). That is, the execution of the overtaking strategy will make the conflict of driving routes urgent, and the ego vehicle may collide with the target obstacle vehicle immediately after the execution of the overtaking strategy.

[0094] The above duration can be set according to the performance of the autonomous driving vehicle, and the embodiments of the present disclosure are not limited to this.

[0095] S105: When the target obstacle vehicle has the driving intention of yielding, executing the overtaking strategy.

[0096] It should be noted that when it is predicted that the driving intention of the target obstacle vehicle is to yield, it can be determined that the overtaking strategy can be executed under the premise of ensuring safety. At this time, the overtaking strategy can be executed.

[0097] Therefore, the disclosed embodiment can play a driving game with the target obstacle vehicle on the premise that the self-driving vehicle is overtaking, thereby improving the traffic efficiency of the autonomous driving vehicle.

[0098] In some embodiments, after the overtaking strategy is executed, it can be detected whether there is still a driving route conflict between the target obstacle vehicle and the vehicle. If there is still a driving route conflict between the target obstacle vehicle and the vehicle, the overtaking strategy can be determined again to predict the driving intention of the target obstacle vehicle after the overtaking strategy is executed, and the overtaking strategy can be executed when the driving intention is to yield.

[0099] That is, the above steps S101 to S105 may be repeatedly executed until the conflict between the target obstacle vehicle and the own vehicle's driving path is resolved.

[0100] In some embodiments, when the target obstacle vehicle has a driving intention of not yielding, a yielding strategy may be determined and executed.

[0101] In some embodiments, the yield strategy may be determined based on the first motion information as: controlling the speed of the ego vehicle to a second speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a second lateral distance.

[0102] The second speed is less than the speed of the target obstacle vehicle, and the second lateral distance is greater than or equal to the lateral distance between the current vehicle and the target obstacle vehicle.

[0103] It is understandable that the giving way strategy has a corresponding relationship with the rushing right strategy to a certain extent.

[0104] For example, when giving way to a target obstacle vehicle, the speed of the vehicle can be made lower than that of the target obstacle vehicle, or the lateral distance between the vehicle and the target obstacle vehicle can be maintained or increased.

[0105] Therefore, the yielding strategy may be determined according to the overtaking strategy also determined by the first movement information.

[0106] Exemplarily, when the overtaking strategy determined according to the first motion information is: controlling the speed of the own vehicle to a first speed, the yielding strategy may be: controlling the speed of the own vehicle to a second speed.

[0107] The overtaking strategy determined according to the first motion information is: controlling the lateral distance between the ego vehicle and the target obstacle vehicle to be a first lateral distance, and the yielding strategy may be: controlling the lateral distance between the ego vehicle and the target obstacle vehicle to be a second lateral distance.

[0108] When the overtaking strategy determined according to the first motion information is: controlling the speed of the own vehicle to a first speed, and controlling the lateral distance between the own vehicle and the target obstacle vehicle to a first lateral distance, the yielding strategy may be: controlling the speed of the own vehicle to a second speed, and controlling the lateral distance between the own vehicle and the target obstacle vehicle to a second lateral distance.

[0109] That is, if it is predicted that executing the rushing strategy will have a high probability of conflicting with the target obstacle vehicle, it also means to a certain extent that executing the yielding strategy opposite to the rushing strategy will have a high probability of resolving the conflict with the target obstacle vehicle. Therefore, the yielding strategy can be determined based on the rushing strategy determined in S102.

[0110] In some embodiments, the yield strategy may also be obtained by inputting the second motion information into a pre-trained neural network model, which is not elaborated in the embodiments of the present disclosure.

[0111] In some embodiments, there may be multiple vehicles in the driving environment of the vehicle. In the automatic driving decision-making method provided in the embodiments of the present disclosure, only one target obstacle vehicle needs to be determined at the same time for the overtaking strategy game.

[0112] At this point, the target obstacle vehicle can be determined by:

[0113] Predict the driving routes of all vehicles in the driving environment of the vehicle. Among them, all vehicles in the driving environment of the vehicle can be understood as all vehicles within a specified radius with the vehicle as the center.

[0114] If there are multiple obstacle vehicles among all vehicles whose driving routes conflict with the ego vehicle, the moments when the multiple obstacle vehicles conflict with the ego vehicle are predicted respectively.

[0115] Among multiple obstacle vehicles, the obstacle vehicle that collides with the vehicle earliest is regarded as the target obstacle vehicle.

[0116] Therefore, the disclosed embodiment can take the vehicle that first conflicts with the self-vehicle as the priority bargaining object, thereby improving the safety of autonomous driving.

[0117] For ease of understanding, two driving scenarios, namely, vehicle lane change and merging from a ramp into a main road, will be taken as examples below to explain in detail the implementation of the embodiments of the present disclosure.

[0118] First, please refer to Figure 2 , Figure 2 A schematic diagram of a driving scenario of a vehicle changing lanes in an embodiment of the present disclosure is shown.

[0119] exist Figure 2 In the scenario shown, the ego vehicle (ego) may want to cut into the adjacent lane due to the need to turn or the slow driving of the vehicle in front, and the target obstacle vehicle (obs) is the vehicle behind the ego vehicle in the adjacent lane.

[0120] At this time, the ego vehicle can first determine the rushing strategy according to the first motion information (one or more of position, speed, acceleration, and lane angle) of the target obstacle vehicle.

[0121] For example, the determined rushing strategy may be to shorten the lateral distance between the vehicle and the target obstacle vehicle, so as to cause the target obstacle vehicle to slow down.

[0122] Next, the ego vehicle can combine the second motion information of the target obstacle vehicle (one or more of position, speed, acceleration, and lane angle) to predict the possible response action of the obstacle vehicle after the execution of the overtaking strategy, that is, the driving intention of the obstacle vehicle. For example, after the ego vehicle executes the overtaking strategy, the obstacle vehicle may slow down to avoid it, or it may speed up to pass through it.

[0123] If the target obstacle vehicle is predicted to yield, the ego vehicle can execute the overtaking strategy. Otherwise, the ego vehicle abandons the execution of the overtaking strategy and maintains the original lane unchanged.

[0124] For example, if the driving route conflict between the ego vehicle and the target obstacle vehicle still exists after the ego vehicle successfully executes the rushing strategy, the ego vehicle can continue to execute the above method and continue to play the rushing game with the target obstacle vehicle until the driving route conflict is resolved.

[0125] If the ego vehicle gives up the strategy of rushing, it can try to change lanes again after a certain time interval and re-play the game with the target obstacle vehicle. At this time, the target obstacle vehicle may still be the target obstacle vehicle of the previous game, or it may be another vehicle.

[0126] Next, please refer to Figure 3 , Figure 3 A schematic diagram of a driving scenario of merging from a ramp into a main road in an embodiment of the present disclosure is shown.

[0127] exist Figure 3 In the illustrated scenario, a target obstacle vehicle (obs) traveling on a ramp wishes to cut in front of the ego vehicle (ego) located on the main road.

[0128] Similar to the previous scenario, the ego vehicle can first determine the overtaking strategy based on the first motion information of the target obstacle vehicle (one or more of position, speed, acceleration, and lane angle).

[0129] For example, the determined overtaking strategy may be to accelerate the vehicle to force the target obstacle vehicle to merge into the main road behind the vehicle.

[0130] Similarly, the ego vehicle can combine the second motion information (position, speed, acceleration, lane angle or one or more) of the target obstacle vehicle to predict the response action that the obstacle vehicle may take after the execution of the overtaking strategy, that is, the driving intention of the obstacle vehicle. For example, after the ego vehicle executes the overtaking strategy, the obstacle vehicle may slow down to avoid it, or it may speed up to pass through it.

[0131] If the target obstacle vehicle's driving intention is predicted to yield, the ego vehicle can execute the overtaking strategy. Otherwise, the ego vehicle abandons the execution of the overtaking strategy and slows down to yield to the target obstacle vehicle.

[0132] Based on the same inventive concept, the present disclosure also provides an automatic driving decision-making device, such as the following embodiment. Figure 1 The method embodiment shown is similar to that shown in FIG. 1 , so the implementation of the automatic driving decision-making device embodiment can refer to the above Figure 1 The implementation of the illustrated method embodiment will not be described in detail in detail.

[0133] Figure 4 A schematic diagram of the structure of an automatic driving decision-making device in an embodiment of the present disclosure is shown.

[0134] Specifically, Figure 4 As shown, the automatic driving decision-making device 400 includes: a first acquisition module 401, a determination module 402, a second acquisition module 403, a prediction module 404, and an execution module 405.

[0135] The first acquisition module 401 is used to acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle.

[0136] The determination module 402 is used to determine a rush-to-go strategy according to the first movement information.

[0137] The second acquisition module 403 is used to acquire the second motion information of the target obstacle vehicle.

[0138] The prediction module 404 is used to predict the driving intention of the target obstacle vehicle after the execution of the rushing strategy based on the second motion information.

[0139] The execution module 405 is used to execute the overtaking strategy when the target obstacle vehicle has the driving intention of yielding.

[0140] In some embodiments, the autonomous driving decision-making device 400 also includes a preprocessing module (not shown in the figure), which is used to predict the driving routes of all vehicles in the driving environment of the own vehicle; if there are multiple obstacle vehicles among all vehicles whose driving routes conflict with the own vehicle, the moments when the multiple obstacle vehicles conflict with the own vehicle are predicted respectively; and the obstacle vehicle among the multiple obstacle vehicles that conflicts with the own vehicle the earliest is used as the target obstacle vehicle.

[0141] In some embodiments, the determination module 402 is specifically used to: determine the overtaking strategy based on the first motion information as: controlling the speed of the vehicle to a first speed, and / or controlling the lateral distance between the vehicle and the target obstacle vehicle to a first lateral distance; wherein the first speed is positively correlated with the speed of the target obstacle vehicle, and the first lateral distance is negatively correlated with the longitudinal distance between the vehicle and the target obstacle vehicle.

[0142] In some embodiments, the prediction module 404 is specifically used to: combine the second motion information with the overtaking strategy, input a pre-trained intention prediction model, and obtain the driving intention of the target obstacle vehicle, wherein the intention prediction model is used to characterize the mapping relationship between the second motion information and the driving intention under different overtaking strategies; or in a pre-constructed driving intention table, query the driving intention corresponding to the second motion information and the overtaking strategy, wherein the driving intention table is used to record the mapping relationship between the second motion information and the driving intention under different overtaking strategies.

[0143] In some embodiments, the execution module 405 is further used to determine a yielding strategy and execute the yielding strategy when the target obstacle vehicle has a driving intention of not yielding.

[0144] In some embodiments, the execution module 405 is also used to determine the yield strategy based on the first motion information: controlling the speed of the ego vehicle to a second speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a second lateral distance; wherein the second speed is less than the speed of the target obstacle vehicle, and the second lateral distance is greater than or equal to the lateral distance between the current ego vehicle and the target obstacle vehicle.

[0145] In some embodiments, the autonomous driving decision-making device 400 also includes a judgment module (not shown in the figure), which is used to re-determine the overtaking strategy if there is still a driving route conflict between the target obstacle vehicle and the own vehicle, predict the driving intention of the target obstacle vehicle after the execution of the overtaking strategy, and execute the overtaking strategy when the driving intention is to give way.

[0146] Refer to the following Figure 5 hereinafter, an electronic device 500 capable of implementing the embodiments of the present disclosure is described. Figure 5 The electronic device 500 shown is merely an example and should not bring any limitation to the functions and scope of application of the embodiments of the present disclosure.

[0147] like Figure 5 As shown, the electronic device 500 is in the form of a general-purpose computing device. The components of the electronic device 500 may include, but are not limited to: at least one processor 510, at least one memory 520, and a bus 530 connecting different system components (including the memory 520 and the processor 510).

[0148] The memory stores program codes, which can be executed by the processor 510, so that the processor 510 executes the steps according to various exemplary embodiments of the present disclosure described in the above “Exemplary Method” section of the present disclosure.

[0149] In some embodiments, the processor 510 may further perform the following steps of the above method embodiment:

[0150] Acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle;

[0151] Determine the rushing strategy according to the first movement information;

[0152] Obtaining the second motion information of the target obstacle vehicle;

[0153] Based on the second motion information, predict the driving intention of the target obstacle vehicle after the execution of the rush strategy;

[0154] When the target obstacle vehicle's driving intention is to yield, the rushing strategy is executed.

[0155] The memory 520 may include a readable medium in the form of a volatile memory, such as a random access memory (RAM) 5201 and / or a cache memory 5202 , and may further include a read-only memory (ROM) 5203 .

[0156] The memory 520 may also include a program / utility 5204 having a set (at least one) of program modules 5205, such program modules 5205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0157] Bus 530 may be a bus representing one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures.

[0158] The electronic device 500 may also communicate with one or more external devices 540 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), one or more devices that enable a user to interact with the electronic device 500, and / or any device that enables the electronic device 500 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed through an input / output (I / O) interface 550. Furthermore, the electronic device 500 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) through a network adapter 560. Figure 5 As shown, the network adapter 560 communicates with other modules of the electronic device 500 via the bus 530. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 500, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0159] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0160] In an exemplary embodiment of the present disclosure, a computer-readable storage medium is also provided, which may be a readable signal medium or a readable storage medium. A program product capable of implementing the above-mentioned method of the present disclosure is stored thereon. In some possible implementations, various aspects of the present disclosure may also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to cause the terminal device to execute the steps according to various exemplary implementations of the present disclosure described in the above-mentioned "Exemplary Method" section of the present disclosure.

[0161] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. An automatic driving decision-making method, characterized in that: include: Acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle; Determining a rush-to-passenger strategy according to the first movement information; Acquiring second motion information of the target obstacle vehicle; Based on the second motion information, predicting the driving intention of the target obstacle vehicle after the overtaking strategy is executed; When the target obstacle vehicle has the driving intention of yielding, the overtaking strategy is executed.

2. The method according to claim 1, characterized in that The method further comprises: Predicting the driving routes of all vehicles in the driving environment of the self-vehicle; If there are multiple obstacle vehicles among all the vehicles whose driving routes conflict with the own vehicle, predict the moments when the multiple obstacle vehicles conflict with the own vehicle respectively; The obstacle vehicle which collides with the own vehicle earliest among the plurality of obstacle vehicles is used as the target obstacle vehicle.

3. The method according to claim 1, characterized in that The step of determining the rushing strategy according to the first movement information includes: According to the first motion information, the overtaking strategy is determined as: controlling the speed of the ego vehicle to a first speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a first lateral distance; The first speed is positively correlated with the speed of the target obstacle vehicle, and the first lateral distance is negatively correlated with the longitudinal distance between the ego vehicle and the target obstacle vehicle.

4. The method according to claim 1, characterized in that The predicting, based on the second motion information, the driving intention of the target obstacle vehicle after the execution of the rushing strategy includes: The second motion information is combined with the overtaking strategy and input into a pre-trained intention prediction model to obtain the driving intention of the target obstacle vehicle, wherein the intention prediction model is used to characterize the mapping relationship between the second motion information and the driving intention under different overtaking strategies; or In a pre-constructed driving intention table, the driving intention corresponding to the second motion information and the overtaking strategy is queried, wherein the driving intention table is used to record the mapping relationship between the second motion information and the driving intention under different overtaking strategies.

5. The method according to claim 1 or 3, characterized in that: The method further comprises: In the case where the target obstacle vehicle has a driving intention of not yielding, determining a yielding strategy; The yield strategy is executed.

6. The method according to claim 5, characterized in that The determining of the yielding strategy includes: According to the first motion information, the yielding strategy is determined as: controlling the speed of the ego vehicle to a second speed, and / or controlling the lateral distance between the ego vehicle and the target obstacle vehicle to a second lateral distance; The second speed is less than the speed of the target obstacle vehicle, and the second lateral distance is greater than or equal to the current lateral distance between the ego vehicle and the target obstacle vehicle.

7. The method according to claim 1, characterized in that After executing the rush-to-go strategy, the method further includes: If there is still a driving route conflict between the target obstacle vehicle and the self-vehicle, the overtaking strategy is determined again, the driving intention of the target obstacle vehicle after the overtaking strategy is executed is predicted, and the overtaking strategy is executed when the driving intention is to give way.

8. An automatic driving decision-making device, characterized in that: include: A first acquisition module is used to acquire first motion information of a target obstacle vehicle, wherein there is a driving route conflict between the target obstacle vehicle and the self-vehicle; A determination module, configured to determine a preemptive strategy according to the first motion information, wherein the preemptive strategy includes at least one of the following: increasing the speed of the ego vehicle, shortening the lateral distance between the ego vehicle and the target obstacle vehicle; A second acquisition module, used to acquire second motion information of the target obstacle vehicle; A prediction module, configured to predict the driving intention of the target obstacle vehicle after the execution of the rushing strategy based on the second motion information; The execution module is used to execute the overtaking strategy when the target obstacle vehicle has the driving intention of giving way.

9. An electronic device, characterized in that: include: processor; as well as A memory, configured to store executable instructions of the processor; The processor is configured to perform the method of any one of claims 1 to 7 by executing the executable instructions.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images