Longitudinal and lateral decision method and device of vehicle and automatic driving vehicle

Abstract

The disclosure discloses a longitudinal and lateral decision method and device of a vehicle and an automatic driving vehicle, relates to the technical field of artificial intelligence, and particularly relates to the technical field of automatic driving. The specific implementation scheme is: determining a dynamic interaction scene of the current vehicle, determining an interaction vehicle interacting with the current vehicle in the dynamic interaction scene, obtaining a driving state of the interaction vehicle, obtaining a vehicle interaction decision model corresponding to the interaction vehicle according to the driving state of the interaction vehicle, and generating a longitudinal and lateral decision result of the current vehicle according to the vehicle interaction decision model corresponding to the interaction vehicle. Therefore, the problem of longitudinal and lateral decision fragmentation in the related art can be solved, and the current vehicle can process various uncertainty problems such as dynamic uncertainty of human driving vehicle behavior and dynamic uncertainty of vehicle behavior interaction in a complex dynamic interaction scene, thereby ensuring the driving safety of the current vehicle to the interaction vehicle.

Description

Technical Field

[0001] This disclosure relates to the field of artificial intelligence technology, and more particularly to the field of autonomous driving technology, specifically to a method, apparatus, and autonomous driving vehicle for making longitudinal and lateral decision-making. Background Technology

[0002] As autonomous driving technology continues to advance and be industrialized in the automotive industry, autonomous vehicles are facing increasingly severe challenges in real traffic environments. Among these challenges, making behavioral decisions for autonomous vehicles in complex and dynamic interactive scenarios is a problem that urgently needs to be solved to achieve advanced autonomous driving and is also a key technology to ensure the safe operation of autonomous vehicles.

[0003] In related technologies, on the one hand, vehicle decision-making methods often adopt a hierarchical approach, separating longitudinal and lateral motion decisions for modular processing. While this approach is beneficial for practical development, in dynamic interactive scenarios, the high coupling between longitudinal and lateral movements means that vehicle safety decisions rely on coordinated longitudinal and lateral decisions (considering both simultaneously), which is detrimental to vehicle safety. On the other hand, existing joint longitudinal and lateral decision-making methods for vehicles face challenges in complex dynamic interactive scenarios. These challenges arise from the highly dynamic and unpredictable driving environment, the significant uncertainty in the behavior of surrounding drivers, and the vehicle's limited understanding of the environment. These challenges include difficulties in coordinating longitudinal and lateral decisions, high uncertainty in the motion of interacting vehicles, and complex modeling of inter-vehicle interaction processes. Summary of the Invention

[0004] This disclosure provides a method, apparatus, and autonomous vehicle for making longitudinal and lateral decisions.

[0005] According to a first aspect of this disclosure, a method for making longitudinal and lateral decisions for a vehicle is provided, comprising:

[0006] Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario.

[0007] Obtain the driving status of the interactive vehicle;

[0008] The vehicle interaction decision model corresponding to the interactive vehicle is obtained based on the driving status of the interactive vehicle.

[0009] The longitudinal and lateral decision results of the current vehicle are generated based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0010] According to a second aspect of this disclosure, a longitudinal and lateral decision-making device for a vehicle is provided, comprising:

[0011] The first determining module is used to determine the dynamic interaction scenario of the current vehicle and to determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario.

[0012] The first acquisition module is used to acquire the driving status of the interactive vehicle;

[0013] The second acquisition module is used to acquire the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle.

[0014] The generation module is used to generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0015] According to a third aspect of this disclosure, an electronic device is provided, comprising:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the longitudinal and lateral decision-making method for the vehicle as described in the first aspect.

[0019] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are used to cause the computer to perform the longitudinal and lateral decision-making method for a vehicle as described in the first aspect.

[0020] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the longitudinal and lateral decision-making method for a vehicle as described in the first aspect.

[0021] According to a sixth aspect of this disclosure, an autonomous vehicle is provided, including electronic devices as described in the third aspect.

[0022] The technical solutions provided in this disclosure have the following beneficial effects:

[0023] By determining the current vehicle's dynamic interaction scenario and identifying the interacting vehicles within that scenario, the driving state of the interacting vehicles is obtained. Based on this driving state, a corresponding vehicle interaction decision model is derived, and then the current vehicle's longitudinal and lateral decision-making results are generated based on this model. This addresses the problem of fragmented longitudinal and lateral decision-making in related technologies, enabling the current vehicle to handle various uncertainties in complex dynamic interaction scenarios, such as the dynamic uncertainty of human driving behavior and the dynamic uncertainty of inter-vehicle interactions, thus ensuring the driving safety of the current vehicle in relation to the interacting vehicles.

[0024] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0025] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein:

[0026] Figure 1 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the first embodiment of this disclosure;

[0027] Figure 2 This is a schematic diagram of a road scene according to the second embodiment of this disclosure;

[0028] Figure 3 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the third embodiment of this disclosure;

[0029] Figure 4 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the fourth embodiment of this disclosure;

[0030] Figure 5 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the fifth embodiment of this disclosure;

[0031] Figure 6 This is a flowchart illustrating the longitudinal and lateral decision-making method for a vehicle provided according to the sixth embodiment of this disclosure;

[0032] Figure 7 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the seventh embodiment of this disclosure;

[0033] Figure 8 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the eighth embodiment of this disclosure;

[0034] Figure 9This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the ninth embodiment of this disclosure;

[0035] Figure 10 This is a schematic diagram illustrating the principle of a vehicle longitudinal and lateral decision-making method in a specific scenario, according to the tenth embodiment of this disclosure.

[0036] Figure 11 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the eleventh embodiment of this disclosure;

[0037] Figure 12 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the twelfth embodiment of this disclosure;

[0038] Figure 13 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the thirteenth embodiment of this disclosure;

[0039] Figure 14 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the fourteenth embodiment of this disclosure;

[0040] Figure 15 A schematic block diagram of an example electronic device 1500 that can be used to implement embodiments of the present disclosure is shown. Detailed Implementation

[0041] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0042] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein.

[0043] In this disclosure, the following embodiments address the following issues in the related technologies: on the one hand, the hierarchical decision-making approach separates the longitudinal and lateral motion decisions of a vehicle, which is not conducive to vehicle safety decisions; on the other hand, existing joint longitudinal and lateral decision-making methods for vehicles face challenges such as difficulty in coordinating longitudinal and lateral decisions, high uncertainty in the motion of interacting vehicles, and complex modeling of the interaction process between vehicles when facing complex dynamic interaction scenarios. Therefore, this disclosure proposes a longitudinal and lateral decision-making method for vehicles.

[0044] The following description, with reference to the accompanying drawings, describes a method, apparatus, and autonomous vehicle for making longitudinal and lateral decisions regarding a vehicle according to embodiments of the present disclosure.

[0045] Figure 1 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the first embodiment of this disclosure.

[0046] It should be noted that the vehicle longitudinal and lateral decision-making method of this disclosure embodiment can be executed by the vehicle longitudinal and lateral decision-making device provided in this disclosure embodiment. This vehicle longitudinal and lateral decision-making device can be an electronic device or can be configured in an electronic device to generate the longitudinal and lateral decision-making result of the current vehicle based on the vehicle interaction decision-making model corresponding to the interactive vehicle interacting with the current vehicle in a dynamic interaction scenario. This solves the problem of fragmented longitudinal and lateral decision-making in related technologies, while enabling the current vehicle to handle various uncertainties such as the dynamic uncertainty of human driving vehicle behavior and the dynamic uncertainty of inter-vehicle behavioral interactions in complex dynamic interaction scenarios, thus ensuring the driving safety of the current vehicle relative to the interactive vehicle.

[0047] The electronic device can be any stationary or mobile computing device capable of data processing, such as a laptop, smartphone, wearable device, or other mobile computing device, or a desktop computer or other stationary computing device, or a server, or other types of computing devices. This embodiment does not impose any restrictions on this.

[0048] like Figure 1 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0049] Step 101: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0050] In this embodiment, the dynamic interaction scenario can be understood as a scenario in which the current vehicle interacts with other vehicles on its own road, specifically including diagonal merging scenarios, horizontal merging scenarios, etc.

[0051] In one possible implementation of this embodiment, the dynamic interaction scenario of the current vehicle can be determined based on whether the road structure of the road where the vehicle is currently located meets the road characteristics under the dynamic interaction scenario. For example, for... Figure 2 The road scene shown is assumed to be vehicle A in this embodiment. Since the road structure of the road where vehicle A is located meets the road characteristics of a diagonal merging scenario, it can be determined that the dynamic interaction scenario of the current vehicle is a diagonal merging scenario.

[0052] In this embodiment, after determining the dynamic interaction scenario of the current vehicle, the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario can be determined.

[0053] In one possible implementation of this embodiment, the interactive vehicle interacting with the current vehicle in the dynamic interaction scene can be determined based on the positional relationship between the current vehicle and other vehicles in the dynamic interaction scene. Optionally, the vehicle located in front of the current vehicle and closest to it in the dynamic interaction scene can be determined as the interactive vehicle interacting with the current vehicle in the dynamic interaction scene. Furthermore, when the dynamic interaction scene is a diagonal merging scene, it is easy to determine that the interaction lane is a diagonal merging lane, thus the vehicle closest to the merging entrance in the diagonal merging lane can be directly determined as the interactive vehicle interacting with the current vehicle in the dynamic interaction scene. For example, for... Figure 2 In the road scenario shown, assuming vehicle A is the current vehicle in this embodiment, on the one hand, since the dynamic interaction scenario is a diagonal merging scenario, it is easy to determine that the interaction lane is the diagonal merging lane, and only vehicle B is located in the diagonal merging lane. Therefore, it can be determined that vehicle B is the interactive vehicle that interacts with vehicle A in this dynamic interaction scenario. On the other hand, although vehicles B, C, and D are all located in front of vehicle A, the relative longitudinal distance ΔD between vehicle B and vehicle A is significant. ab The distance between vehicle C and vehicle A is smaller than the distance between vehicle D and vehicle A. Therefore, it can be determined that vehicle B is closest to vehicle A, and thus vehicle B is the interactive vehicle that interacts with vehicle A in the dynamic interactive scenario.

[0054] Step 102: Obtain the driving status of the interactive vehicle.

[0055] In this embodiment, the driving state of the interactive vehicle can be obtained. It should be noted that the driving state of the interactive vehicle obtained in this step is mainly used to distinguish whether the interactive vehicle is driving rationally. It is only a priori judgment to facilitate the selection of subsequent decision-making models and cannot fully represent the actual driving state of the interactive vehicle.

[0056] Since the driving status of the interactive vehicle obtained in this step is mainly used to distinguish whether the interactive vehicle is driving rationally, in this embodiment, the driving status of the interactive vehicle can include rational driving status and irrational driving status.

[0057] Step 103: Obtain the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle.

[0058] In this embodiment, after obtaining the driving state of the interactive vehicle, a vehicle interaction decision model corresponding to the interactive vehicle can be obtained based on the driving state. Specifically, when the driving state of the interactive vehicle is a rational driving state, the vehicle interaction decision model corresponding to the interactive vehicle in the rational driving state can be obtained; when the driving state of the interactive vehicle is an irrational driving state, the vehicle interaction decision model corresponding to the interactive vehicle in the irrational driving state can be obtained.

[0059] Step 104: Generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0060] In this embodiment, the longitudinal and lateral decision results of the current vehicle can be understood as the longitudinal and lateral behavior strategies of the current vehicle, specifically whether the current vehicle yields in the longitudinal direction and avoids in the lateral direction.

[0061] In this embodiment, the longitudinal and lateral decision results of the current vehicle can be generated based on the vehicle interaction decision model corresponding to the interactive vehicle. The longitudinal and lateral decision results can include longitudinal yielding and lateral avoidance, longitudinal yielding and lateral non-yielding, longitudinal non-yielding and lateral avoidance, and longitudinal non-yielding and lateral non-yielding.

[0062] The longitudinal and lateral decision-making method for vehicles provided in this disclosure determines the dynamic interaction scenario of the current vehicle and the interactive vehicles interacting with the current vehicle in the dynamic interaction scenario, thereby obtaining the driving state of the interactive vehicles. Based on the driving state of the interactive vehicles, a vehicle interaction decision model corresponding to the interactive vehicles is obtained, and then the longitudinal and lateral decision-making results of the current vehicle are generated based on the vehicle interaction decision model corresponding to the interactive vehicles. Therefore, while solving the problem of fragmented longitudinal and lateral decision-making in related technologies, it enables the current vehicle to handle various uncertainties in complex dynamic interaction scenarios, such as the dynamic uncertainty of human driving behavior and the dynamic uncertainty of inter-vehicle interaction, ensuring the driving safety of the current vehicle in relation to the interactive vehicles.

[0063] As can be seen from the above analysis, in this embodiment of the disclosure, the dynamic interaction scenario of the current vehicle and the interactive vehicles interacting with the current vehicle in the dynamic interaction scenario can be determined. To clearly explain how the dynamic interaction scenario and interactive vehicles are determined in this disclosure, this embodiment of the disclosure provides another method for vehicle longitudinal and lateral decision-making. Figure 3 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the third embodiment of this disclosure.

[0064] like Figure 3 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0065] Step 301: Obtain the road image captured by the current vehicle.

[0066] In this embodiment, the vehicle is equipped with image acquisition devices, such as a dashcam or a vehicle camera, so that the vehicle can acquire road images of the road it is on based on the image acquisition devices it is equipped with.

[0067] In this embodiment, the vehicle's longitudinal and lateral decision-making device can acquire the road images collected by the vehicle in various open, legal, and compliant ways. For example, the vehicle's longitudinal and lateral decision-making device can acquire the road images collected by the vehicle in real time during the process of the vehicle collecting road images, or it can acquire the road images collected by the vehicle from other devices through network transmission or physical copying, or it can acquire the road images collected by the vehicle in other open, legal, and compliant ways. This embodiment does not impose any restrictions on this.

[0068] Step 302: Obtain road features from the road image and determine the dynamic interaction scenario of the current vehicle based on the road features.

[0069] In this embodiment, the dynamic interaction scenario can be understood as a scenario in which the current vehicle interacts with other vehicles on its own road, specifically including diagonal merging scenarios, horizontal merging scenarios, etc.

[0070] In this embodiment, road features can be obtained based on road images captured by the current vehicle, thereby determining the dynamic interaction scenario of the current vehicle based on the road features. For example, it can be determined whether the obtained road features meet the requirements of a diagonal merging scenario. If the road features meet the requirements of a diagonal merging scenario, the dynamic interaction scenario of the current vehicle is determined to be a diagonal merging scenario; otherwise, if the road features do not meet the requirements of a diagonal merging scenario, the vehicle will proceed to the determination of other dynamic interaction scenarios.

[0071] Step 303: Obtain other vehicles in the dynamic interactive scene.

[0072] In this embodiment, after determining the dynamic interaction scenario, other vehicles besides the current vehicle in the dynamic interaction scenario can be acquired. Optionally, other vehicles in the dynamic interaction scenario can be acquired based on road images captured by the current vehicle.

[0073] It should be noted that the number of other vehicles in the dynamic interaction scene obtained in this step can be one or more, and this embodiment does not impose any restrictions on this.

[0074] Step 304: Calculate the distance between the current vehicle and other vehicles in the dynamic interaction scene.

[0075] In this embodiment, the distance between the current vehicle and other vehicles in the dynamic interaction scene can be calculated to determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scene based on the distance between the current vehicle and other vehicles in the dynamic interaction scene.

[0076] It should be noted that in dynamic interaction scenarios, other vehicles are those merging diagonally into the lane, such as... Figure 2 When vehicle B is in the road scene shown, calculating the distance between the current vehicle and other vehicles in the dynamic interaction scene is equivalent to calculating the relative longitudinal distance between the current vehicle and vehicles merging diagonally into the lane.

[0077] Step 305: Select the vehicle that is closest to the current vehicle among the other vehicles as the interaction vehicle.

[0078] In this embodiment, after calculating the distance between the current vehicle and other vehicles in the dynamic interaction scene, the vehicle that is closest to the current vehicle among the other vehicles can be selected as the interactive vehicle that interacts with the current vehicle in the dynamic interaction scene based on the distance between the current vehicle and other vehicles in the dynamic interaction scene.

[0079] Step 306: Obtain the driving status of the interactive vehicle.

[0080] Step 307: Obtain the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle.

[0081] Step 308: Generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0082] It should be noted that the specific implementation process of steps 306-308 can be found in the description of steps 102-104 in the previous embodiment, and will not be repeated here.

[0083] The dialogue processing method provided in this disclosure acquires road images captured by the current vehicle, obtains road characteristics based on the road images, determines the dynamic interaction scene of the current vehicle based on the road characteristics, and calculates the distance between the current vehicle and other vehicles in the dynamic interaction scene by acquiring other vehicles in the dynamic interaction scene. The vehicle closest to the current vehicle among the other vehicles is then identified as the interactive vehicle. Thus, it is possible to determine the dynamic interaction scene of the current vehicle based on prop characteristics, and to determine the interactive vehicle interacting with the current vehicle in the dynamic interaction scene based on the distance between vehicles.

[0084] As can be seen from the above analysis, the driving state of the interactive vehicle can be obtained in this embodiment of the disclosure. To clearly explain how the driving state of the interactive vehicle is obtained in this disclosure, this embodiment of the disclosure provides another method for longitudinal and lateral decision-making of the vehicle. Figure 4 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the fourth embodiment of this disclosure.

[0085] like Figure 4 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0086] Step 401: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0087] It should be noted that the specific implementation process of this step can be found in the description of step 101 in the above embodiment, and will not be repeated here.

[0088] Step 402: Obtain the merging endpoint of the road where the interactive vehicle is located.

[0089] It should be noted that, in this embodiment, the dynamic interaction scenario is as follows: Figure 2 The diagonal merging scenario shown allows for the determination of the merging endpoint of the road where the interactive vehicle is located after the dynamic interaction scenario and the interactive vehicle have been determined. Optionally, if the current vehicle is equipped with an image acquisition device and has acquired a road image of its location based on the image acquisition device, the merging endpoint of the road where the interactive vehicle is located can be obtained based on the road image acquired by the current vehicle.

[0090] Step 403: Obtain the collision distance G of the interactive vehicle relative to the merging endpoint.

[0091] In this embodiment, after obtaining the merging endpoint of the road where the interactive vehicle is located, the relative longitudinal distance ΔD1 between the interactive vehicle and the merging endpoint can be calculated. After obtaining the longitudinal velocity v1 of the interactive vehicle, the head-on collision distance G of the interactive vehicle relative to the merging endpoint can be calculated using the formula G=ΔD1v1 based on the relative longitudinal distance ΔD1 between the interactive vehicle and the merging endpoint and the longitudinal velocity v1 of the interactive vehicle.

[0092] Step 404: Obtain the G sequence of the interactive vehicle within a preset time period, and cluster the G sequence to generate clustering results.

[0093] In this embodiment, after obtaining the head-on collision distance G of the interactive vehicle relative to the merging endpoint, the G sequence of the interactive vehicle within a preset time period can be obtained, and the G sequence can be clustered using sequence clustering to generate clustering results. The specific setting of the preset time period is not limited in this embodiment. Optionally, it can be set based on human experience; for example, the preset time period can be set to 10:00-10:10. Alternatively, it can be dynamically adjusted according to actual application needs. This embodiment does not impose any restrictions on this.

[0094] In this embodiment, the sequence clustering method is used to cluster the G sequence, and the generated clustering results represent rational driving state and irrational driving state respectively.

[0095] Step 405: Obtain the driving status of the interactive vehicle based on the clustering results.

[0096] Since the clustering results generated in the previous step represent rational and irrational driving states respectively, the driving state of the interactive vehicle can be obtained based on the clustering results.

[0097] Step 406: Obtain the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle.

[0098] Step 407: Generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0099] It should be noted that the specific implementation process of steps 406-407 can be found in the description of steps 103-104 of the above embodiment, and will not be repeated here.

[0100] The longitudinal and lateral decision-making method for vehicles provided in this disclosure obtains the merging endpoint of the road where the interactive vehicle is located, thereby obtaining the head-on collision distance G of the interactive vehicle relative to the merging endpoint. This allows for the acquisition of the G sequence of the interactive vehicle within a preset time period. The G sequence is then clustered to generate clustering results, and the driving state of the interactive vehicle is obtained based on the clustering results. Therefore, by using sequence clustering to cluster the G sequence of the interactive vehicle within a preset time period, the driving state of the interactive vehicle can be obtained based on the clustering results.

[0101] In this embodiment of the disclosure, after determining the interacting vehicles, the head-on collision time distance between the current vehicle and the interacting vehicle can be obtained. Based on the head-on collision time distance between the current vehicle and the interacting vehicle, the degree of interaction between the current vehicle and the interacting vehicle can be determined. To clearly illustrate this process, this embodiment of the disclosure provides another method for vehicle longitudinal and lateral decision-making. Figure 5 This is a schematic flowchart of the longitudinal and lateral decision-making method for vehicles provided according to the fifth embodiment of this disclosure.

[0102] like Figure 5 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0103] Step 501: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0104] It should be noted that the specific implementation process of this step can be found in the description of step 101 in the above embodiment, and will not be repeated here.

[0105] Step 502: Obtain the head-on collision time distance between the current vehicle and the interacting vehicle.

[0106] In this embodiment, after determining the dynamic interaction scenario and the interactive vehicle, the head-on collision time distance between the current vehicle and the interactive vehicle can be obtained. This head-on collision time distance can be calculated using the relative longitudinal distance between the current vehicle and the interactive vehicle, and the longitudinal speed of the vehicle behind. For example, for... Figure 2 In the diagonally merging scenario shown, the current vehicle is vehicle A, and the interacting vehicle is vehicle B. The head-on collision distance between vehicle A and vehicle B is W = ΔD. ab v rear , where ΔD ab v represents the relative longitudinal distance between vehicle A and vehicle B. rear This indicates the longitudinal speed of the vehicle located behind.

[0107] Step 503: If the head-on collision distance is less than the first threshold, then the vehicle interaction decision model corresponding to the interactive vehicle is further obtained based on the driving state of the interactive vehicle.

[0108] In this embodiment, if the head-on collision distance between the current vehicle and the interacting vehicle is less than a first threshold, it indicates that the interaction between the current vehicle and the interacting vehicle is strong. Further acquisition of the driving state of the interacting vehicle is required, and a vehicle interaction decision model corresponding to the interacting vehicle is obtained based on the driving state. The specific value of the first threshold is not limited in this embodiment. Optionally, it can be set based on human experience; for example, the first threshold can be set to 50. Alternatively, it can be dynamically adjusted according to actual application needs. This embodiment does not impose any restrictions on this.

[0109] Step 504: Generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0110] It should be noted that the specific implementation process of this step can be found in the description of step 104 of the above embodiment, and will not be repeated here.

[0111] Step 505: If the distance to the frontal collision is greater than the second threshold, then stop the decision-making process, where the second threshold is greater than the first threshold.

[0112] In this embodiment, if the head-on collision distance between the current vehicle and the interacting vehicle is greater than the second threshold, it indicates that the interaction between the current vehicle and the interacting vehicle is weak. In this case, no decision needs to be made, and the decision-making process can be stopped. The second threshold must be greater than the first threshold, but this embodiment does not limit the specific value of the second threshold. Optionally, it can be set based on human experience. For example, if the first threshold is set to 50, the second threshold can be set to any data greater than 50, such as 70, 80, or 100. Alternatively, it can be dynamically adjusted according to actual application needs. This embodiment does not impose any restrictions on this.

[0113] Step 506: If the distance to the frontal collision is greater than or equal to the first threshold and less than or equal to the second threshold, then the longitudinal and lateral decision results of the previous frame are used as the longitudinal and lateral decision results of the vehicle in the current frame.

[0114] In this embodiment, if the head-on collision time distance between the current vehicle and the interacting vehicle is greater than or equal to the first threshold and less than or equal to the second threshold, that is, the head-on collision time distance between the current vehicle and the interacting vehicle falls within the interval of [first threshold, second threshold], then the longitudinal and lateral decision results of the previous frame can be used as the longitudinal and lateral decision results of the vehicle in the current frame. Thus, the consistency of the results between frames can be guaranteed.

[0115] The longitudinal and lateral decision-making method for vehicles provided in this disclosure obtains the head-on collision time distance between the current vehicle and the interacting vehicle. If the head-on collision time distance is less than a first threshold, the method further obtains the vehicle interaction decision model corresponding to the interacting vehicle based on its driving state, and generates the longitudinal and lateral decision result for the current vehicle based on this model. Alternatively, if the head-on collision time distance is greater than a second threshold (where the second threshold is greater than the first threshold), the decision-making process stops. Or, if the head-on collision time distance is greater than or equal to the first threshold and less than or equal to the second threshold, the longitudinal and lateral decision result of the previous frame is used as the longitudinal and lateral decision result for the current frame. Therefore, different decisions can be made based on different levels of interaction between the current vehicle and the interacting vehicle.

[0116] To clearly illustrate the process of generating the longitudinal and lateral decision-making results of the current vehicle based on the vehicle interaction decision-making model corresponding to the interactive vehicle in the above embodiments, this disclosure provides another method for vehicle longitudinal and lateral decision-making. Figure 6 This is a schematic flowchart of the longitudinal and lateral decision-making method for vehicles provided according to the sixth embodiment of this disclosure.

[0117] like Figure 6 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0118] Step 601: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0119] It should be noted that the specific implementation process of this step can be found in the description of step 101 in the above embodiment, and will not be repeated here.

[0120] Step 602: Obtain the driving status of the interactive vehicle.

[0121] In this embodiment, the driving state of the interactive vehicle is a rational driving state.

[0122] The other implementation process of this step can be found in the description of step 102 of the above embodiment, and will not be repeated here.

[0123] Step 603: Obtain the road to which the current vehicle belongs, and the road to which the interacting vehicle belongs.

[0124] Because different right-of-way conditions affect a vehicle's decision-making process—for example, the actions a vehicle takes on roads with different right-of-way conditions may differ—it is necessary to determine the right-of-way relationship between the current vehicle and interacting vehicles in advance. Right-of-way refers to the rights of traffic participants, which are their rights to engage in road traffic activities within a certain space and time, as stipulated by traffic regulations.

[0125] In this embodiment, after obtaining the driving status of the interactive vehicle, the road to which the current vehicle belongs and the road to which the interactive vehicle belongs can be obtained to determine the right-of-way relationship between the current vehicle and the interactive vehicle.

[0126] Step 604: Determine the right-of-way relationship between the current vehicle and the interacting vehicle based on the road to which the current vehicle belongs and the road to which the interacting vehicle belongs.

[0127] In this embodiment, the right-of-way relationship between the current vehicle and the interacting vehicle can be understood as the relative right-of-way priority of the current vehicle compared to the interacting vehicle. Optionally, the right-of-way relationship between the current vehicle and the interacting vehicle may include: the current vehicle having high right-of-way priority and the interacting vehicle having low right-of-way priority, or the current vehicle having low right-of-way priority and the interacting vehicle having high right-of-way priority, etc.

[0128] In this embodiment, after obtaining the road to which the current vehicle belongs and the road to which the interacting vehicle belongs, the right-of-way relationship between the current vehicle and the interacting vehicle can be determined based on these roads. Optionally, the right-of-way relative to the interacting vehicle can be determined according to traffic regulations based on the roads to which the current vehicle belongs and the roads to which the interacting vehicle belongs. For example, for... Figure 2 In the road scenario shown, according to traffic regulations, vehicle A (the current vehicle) has high right-of-way, while vehicle B (the interacting vehicle) has low right-of-way.

[0129] It should be noted that in this embodiment, a lane change decision can be initiated when the current vehicle has low right-of-way. In other words, the current vehicle can initiate a lane change decision when its right-of-way is low. Therefore, if the current vehicle initiates a lane change decision and the lane change is successful, no further decision is required, and the decision-making process can be stopped. If the current vehicle initiates a lane change decision but the lane change fails, further decision-making is required.

[0130] Step 605: Obtain the interaction space between the current vehicle and the interacting vehicle.

[0131] In this embodiment, since the driving state of the interactive vehicle is a rational driving state, the vehicle interaction decision model corresponding to the interactive vehicle in the rational driving state can be obtained, and the vehicle interaction decision stage in the rational driving state can be entered.

[0132] In this embodiment, the interaction space between the current vehicle and the interacting vehicle can be obtained first. Optionally, the interaction space between the current vehicle and the interacting vehicle... It can include four types: longitudinal yielding and lateral yielding, longitudinal yielding without lateral yielding, longitudinal not yielding and lateral yielding, and longitudinal not yielding and lateral not yielding.

[0133] Step 606: Generate the first cost function for the current vehicle and the second cost function for the interacting vehicle, respectively.

[0134] In this embodiment, after obtaining the interaction space between the current vehicle and the interacting vehicle, a first cost function for the current vehicle and a second cost function for the interacting vehicle can be generated respectively. Optionally, the first cost function C for the current vehicle can be generated based on the vehicle's safety and efficiency. a The second cost function C of the interactive vehicle b .

[0135] Step 607: Adjust the first weight factor of the first cost function according to the right-of-way relationship, and set the second weight factor of the second cost function as a priori value.

[0136] In this embodiment, the first weight factor of the first cost function can be adjusted according to the right-of-way relationship between the current vehicle and the interacting vehicle, and the second weight factor of the second cost function can be set as a priori value. Optionally, the first cost function C can be adjusted based on the current vehicle's right-of-way level. a The first weighting factor λ a and the second cost function C b The second weighting factor λ b Set as a priori value.

[0137] Step 608: Generate the longitudinal and lateral decision results of the current vehicle based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor.

[0138] In this embodiment, the longitudinal and lateral decision-making results of the current vehicle can be generated based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor. Optionally, the equilibrium solution method of a game theory model can be used to obtain the interaction action space between the current vehicle and the interacting vehicles. The probability values ​​of different interactive actions are calculated, and the interactive action corresponding to the highest probability value is taken as the output to obtain the longitudinal and lateral decision results of the current vehicle and the longitudinal and lateral decision results of the interacting vehicle.

[0139] The longitudinal and lateral decision-making method for vehicles provided in this disclosure obtains the road to which the current vehicle belongs and the road to which the interacting vehicle belongs. It then determines the right-of-way relationship between the current vehicle and the interacting vehicle based on these two roads. After obtaining the interaction space between the current vehicle and the interacting vehicle, it generates a first cost function for the current vehicle and a second cost function for the interacting vehicle. Based on the right-of-way relationship, it adjusts a first weight factor of the first cost function and sets a second weight factor of the second cost function as a priori value. Finally, it generates the longitudinal and lateral decision-making result for the current vehicle based on the first cost function, the second cost function, the first weight factor, and the second weight factor. Therefore, when the interacting vehicle is in a rational driving state, it can obtain the vehicle interaction decision-making model corresponding to the interacting vehicle based on its driving state and generate the longitudinal and lateral decision-making result for the current vehicle based on the vehicle interaction decision-making model corresponding to the interacting vehicle.

[0140] To ensure high accuracy in generating longitudinal and lateral decision-making results for the current vehicle when the interactive vehicle is in a rational driving state, this disclosure provides another method for longitudinal and lateral decision-making for vehicles. Figure 7 This is a schematic flowchart of the longitudinal and lateral decision-making method for vehicles provided according to the seventh embodiment of this disclosure.

[0141] like Figure 7 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0142] Step 701: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0143] Step 702: Obtain the driving status of the interactive vehicle.

[0144] Step 703: Obtain the road to which the current vehicle belongs, and the road to which the interacting vehicle belongs.

[0145] Step 704: Determine the right-of-way relationship between the current vehicle and the interacting vehicle based on the road to which the current vehicle belongs and the road to which the interacting vehicle belongs.

[0146] Step 705: Obtain the interaction space between the current vehicle and the interacting vehicle.

[0147] Step 706: Generate the first cost function for the current vehicle and the second cost function for the interacting vehicle, respectively.

[0148] Step 707: Adjust the first weight factor of the first cost function according to the right-of-way relationship, and set the second weight factor of the second cost function as a priori value.

[0149] Step 708: Generate the longitudinal and lateral decision results of the current vehicle based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor.

[0150] It should be noted that the specific implementation process of steps 701-708 can be found in the description of steps 601-608 in the above embodiment, and will not be repeated here.

[0151] Step 709: Obtain the actual interactive actions of the interactive vehicle.

[0152] Due to the uncertainty of human drivers, the actual driving behavior of interactive vehicles will differ from the driving behavior preset by the decision model. Therefore, it is necessary to adjust the parameters of the decision model so that the prediction results of the interactive vehicles output by the decision model are consistent with their actual driving results.

[0153] In this embodiment, a feedback correction mechanism of "prior hypothesis - introduction of observation - posterior correction" can be adopted, since the second cost function C of the interactive vehicle has been included. b The second weighting factor κ b This is set as a priori value, allowing the actual interaction actions of the vehicle in the previous frame to be obtained after the decision model starts running, thus determining whether the actual interaction actions of the vehicle in the previous frame belong to the interaction action space. Which of the following situations applies, i.e., when actual observation data is introduced?

[0154] Step 710: Adjust the second weighting factor based on the actual interactive actions of the interactive vehicle and the predicted interactive actions of the interactive vehicle.

[0155] In this embodiment, after introducing actual observation data, posterior correction can be performed, that is, the second weighting factor can be adjusted based on the actual interaction actions of the interactive vehicle and the predicted interaction actions of the interactive vehicle. Optionally, if the actual interaction actions of the interactive vehicle are consistent with the longitudinal and lateral decision results of the interactive vehicle output by the decision model in the previous frame, no correction is made; if they are inconsistent, the second weighting factor κ is adjusted based on the actual interaction actions of the interactive vehicle and the predicted interaction actions of the interactive vehicle. b Adjustments were made to create a new second weighting factor κ. b The longitudinal and lateral decision-making results of the interactive vehicle output by the decision-making model are consistent with the actual interactive actions of the interactive vehicle.

[0156] The longitudinal and lateral decision-making method for vehicles provided in this disclosure adjusts a second weighting factor based on the actual interaction actions of the interacting vehicles and the predicted interaction actions of the interacting vehicles, thereby effectively improving the accuracy of generating the longitudinal and lateral decision-making results for the current vehicle.

[0157] The above embodiments describe the implementation process of the longitudinal and lateral decision-making method for the vehicle when the driving state of the interactive vehicle is rational driving. The following will combine... Figure 8 The implementation process of the longitudinal and lateral decision-making method for interactive vehicles when the driving state is irrational is further explained.

[0158] Figure 8 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the eighth embodiment of this disclosure.

[0159] like Figure 8 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0160] Step 801: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0161] It should be noted that the specific implementation process of this step can be found in the description of step 101 in the above embodiment, and will not be repeated here.

[0162] Step 802: Obtain the driving status of the interactive vehicle.

[0163] In this embodiment, the driving state of the interactive vehicle is an irrational driving state.

[0164] The other implementation process of this step can be found in the description of step 102 of the above embodiment, and will not be repeated here.

[0165] Step 803: Obtain the road to which the current vehicle belongs, and the road to which the interacting vehicle belongs.

[0166] Step 804: Determine the right-of-way relationship between the current vehicle and the interacting vehicle based on the road to which the current vehicle belongs and the road to which the interacting vehicle belongs.

[0167] It should be noted that the specific implementation process of steps 803-804 can be found in the description of steps 603-604 in the above embodiment, and will not be repeated here.

[0168] Step 805: Obtain the first distance of the current vehicle relative to the merging destination, and the second distance of the interacting vehicle relative to the merging destination.

[0169] In this embodiment, since the driving state of the interactive vehicle is irrational driving, the vehicle interaction decision model corresponding to the interactive vehicle in the irrational driving state can be obtained, and the vehicle interaction decision stage in the irrational driving state can be entered.

[0170] In this embodiment, the longitudinal and lateral decision results of the current vehicle can be generated based on the current vehicle's right-of-way and the distances of the current vehicle and interacting vehicles relative to the merging destination.

[0171] In this embodiment, a first distance of the current vehicle relative to the merging destination and a second distance of the interacting vehicle relative to the merging destination can be obtained to determine which vehicle, the current vehicle or the interacting vehicle, is closer to the merging destination based on the first and second distances. It is understood that when the first distance is greater than the second distance, it indicates that the current vehicle is closer to the merging destination than the interacting vehicle, and vice versa.

[0172] Step 806: If the current vehicle is determined to have high right-of-way based on the right-of-way relationship, and the first distance is greater than the second distance, then control the current vehicle to select the longitudinal non-yield and lateral non-avoidance strategy.

[0173] In this embodiment, if the current vehicle is determined to have high right-of-way based on the right-of-way relationship between the current vehicle and the interacting vehicle, and the first distance of the current vehicle relative to the merging endpoint is greater than the second distance of the interacting vehicle relative to the merging endpoint, then the current vehicle can be controlled to select the longitudinal non-yield and lateral non-avoidance strategy.

[0174] Step 807: If the current vehicle is determined to have low right-of-way based on the right-of-way relationship, and the first distance is less than or equal to the second distance, then control the current vehicle to select a longitudinal yielding and lateral avoidance strategy.

[0175] In this embodiment, if the current vehicle is determined to have low right-of-way based on the right-of-way relationship between the current vehicle and the interacting vehicle, and the first distance of the current vehicle relative to the merging endpoint is less than or equal to the second distance of the interacting vehicle relative to the merging endpoint, then the current vehicle can be controlled to select a longitudinal yielding and lateral avoidance strategy.

[0176] The longitudinal and lateral decision-making method for vehicles provided in this disclosure obtains the road to which the current vehicle belongs and the road to which the interacting vehicle belongs. It then determines the right-of-way relationship between the current vehicle and the interacting vehicle based on their respective roads, and obtains a first distance of the current vehicle relative to the merging endpoint and a second distance of the interacting vehicle relative to the merging endpoint. If, based on the right-of-way relationship, the current vehicle is determined to have high right-of-way and the first distance is greater than the second distance, the current vehicle is controlled to choose a longitudinal non-yield and lateral non-avoidance strategy. Alternatively, if, based on the right-of-way relationship, the current vehicle is determined to have low right-of-way and the first distance is less than or equal to the second distance, the current vehicle is controlled to choose a longitudinal yield and lateral avoidance strategy. Therefore, even when the interacting vehicle is in an irrational driving state, the method can obtain the vehicle interaction decision model corresponding to the interacting vehicle based on its driving state and generate the longitudinal and lateral decision results for the current vehicle based on this model.

[0177] The previous embodiment described an implementation of a longitudinal and lateral decision-making method for an interactive vehicle when its driving state is irrational. This embodiment provides another implementation of the same method. Figure 9 This is a flowchart illustrating the longitudinal and lateral decision-making method for vehicles provided according to the ninth embodiment of this disclosure.

[0178] like Figure 9 As shown, the vehicle's longitudinal and lateral decision-making method includes the following steps:

[0179] Step 901: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicles that interact with the current vehicle in the dynamic interaction scenario.

[0180] Step 902: Obtain the driving status of the interactive vehicle.

[0181] It should be noted that the specific implementation process of steps 901-902 can be found in the description of steps 801-802 in the previous embodiment, and will not be repeated here.

[0182] Step 903: Obtain the driving behavior of the interactive vehicle.

[0183] In this embodiment, since the driving state of the interactive vehicle is irrational driving, the vehicle interaction decision model corresponding to the interactive vehicle in the irrational driving state can be obtained, and the vehicle interaction decision stage in the irrational driving state can be entered.

[0184] In this embodiment, the longitudinal and lateral decision results of the current vehicle can be generated based on the driving behavior of the interactive vehicle.

[0185] In this embodiment, since the driving state of the interactive vehicle is irrational driving, the driving behavior of the interactive vehicle obtained is irrational driving behavior, which can specifically include two types: aggressive driving behavior and conservative driving behavior.

[0186] Step 904: If the driving behavior is aggressive driving behavior, then control the current vehicle to select a longitudinal yielding strategy, and select a lateral avoidance strategy based on the relative lateral distance between the current vehicle and the interacting vehicle.

[0187] In this embodiment, if the driving behavior of the interactive vehicle is aggressive driving behavior, it can be assumed that the interactive vehicle will make a longitudinal non-yield decision, thereby controlling the current vehicle to select a longitudinal yielding strategy.

[0188] In this embodiment, the lateral decision of the current vehicle can be determined based on the relative lateral distance between the current vehicle and the interacting vehicle. Optionally, whether to select a lateral avoidance strategy can be determined based on whether the relative lateral distance between the current vehicle and the interacting vehicle is greater than a preset threshold. The specific value of the preset threshold is not limited in this embodiment; it can be set based on human experience, for example, setting the preset threshold to 50, or it can be dynamically adjusted according to actual application needs. This embodiment does not impose any restrictions on this.

[0189] Step 907: If the driving behavior is conservative, then control the current vehicle to select the longitudinal non-yield strategy, and select the lateral avoidance strategy based on the relative lateral distance between the current vehicle and the interacting vehicle.

[0190] In this embodiment, if the driving behavior of the interactive vehicle is conservative, it can be assumed that the interactive vehicle will make a longitudinal yielding decision, thereby controlling the current vehicle to choose a longitudinal non-yielding strategy.

[0191] Similarly, in this embodiment, the lateral decision result of the current vehicle can also be determined based on the relative lateral distance between the current vehicle and the interacting vehicle. Optionally, whether to select a lateral avoidance strategy can be determined based on whether the relative lateral distance between the current vehicle and the interacting vehicle is greater than a preset threshold. The specific value of the preset threshold is not limited in this embodiment. Optionally, it can be set based on human experience, for example, setting the preset threshold to 50, or it can be dynamically adjusted according to actual application needs. This embodiment does not impose any restrictions on this.

[0192] The longitudinal and lateral decision-making method for vehicles provided in this disclosure obtains the road to which the current vehicle belongs and the road to which the interacting vehicle belongs. It then determines the right-of-way relationship between the current vehicle and the interacting vehicle based on their respective roads and obtains the driving behavior of the interacting vehicle. If the driving behavior is aggressive, the current vehicle is controlled to select a longitudinal yielding strategy and a lateral avoidance strategy based on the relative lateral distance between the current vehicle and the interacting vehicle. Alternatively, if the driving behavior is conservative, the current vehicle is controlled to select a longitudinal non-yielding strategy and a lateral avoidance strategy based on the relative lateral distance between the current vehicle and the interacting vehicle. Therefore, even when the interacting vehicle is in an irrational driving state, the method can obtain the corresponding vehicle interaction decision model based on the interacting vehicle's driving state and generate the longitudinal and lateral decision results for the current vehicle based on this model.

[0193] To clearly illustrate the above-disclosed embodiments, examples are provided below.

[0194] Figure 10 This is a schematic diagram illustrating the principle of a vehicle longitudinal and lateral decision-making method in a specific scenario, provided in the tenth embodiment of this disclosure.

[0195] like Figure 10 As shown, the vehicle joint decision-making module mainly includes the following three steps:

[0196] Step 1, Interactive Scene Understanding, mainly includes four sub-steps: determining the dynamic interactive scene, determining the interactive vehicle, judging the degree of vehicle interaction, and judging the motion cognition of the interactive vehicle.

[0197] Step 1.1, determining the dynamic interaction scenario, primarily relies on whether the road structure of the current vehicle's location meets the road characteristics required for a dynamic interaction scenario. For example, it can be determined whether the road structure of the current vehicle's location meets the road characteristics required for a diagonal merging scenario. If it does, proceed to Step 1.2, determining the interacting vehicle; otherwise, proceed to the judgment of other dynamic interaction scenarios. (For...) Figure 2 The road scene shown can be determined to meet the characteristics of a road in a diagonal merging scenario, so it will proceed to step 1.2.

[0198] Step 1.2, determining the interactive vehicle, can be based on the positional relationship between the current vehicle and other vehicles in the dynamic interaction scenario to identify the interactive vehicle that will interact with the current vehicle in the dynamic interaction scenario. For example, in a diagonal merging scenario, the vehicle closest to the merging entrance in the diagonal merging lane can be searched and identified as the interactive vehicle that will interact with the current vehicle in the dynamic interaction scenario. Figure 2 The road scene shown is based on Figure 2 Based on the positions of vehicle A (the current vehicle) and other vehicles, it can be determined that the vehicle interacting with vehicle A is vehicle B.

[0199] Step 1.3, Determining the degree of vehicle interaction. Based on the identified interacting vehicle being vehicle B, analyze the strength of its interaction with vehicle A. Calculate the head-on collision distance W between vehicle A and vehicle B based on vehicle B's current speed:

[0200]

[0201] Where, ΔD ab v represents the relative longitudinal distance between vehicle A and vehicle B. rear This represents the longitudinal speed of the vehicle located behind.

[0202] Define an interval [α, β]. If W is less than α, it is judged as a strong interaction and proceeds to step 1.4; if W is greater than α, it is judged as a weak interaction, and the subsequent steps are not continued, and the decision module is exited; if it falls within this interval, the result of the previous frame is used to ensure the consistency of the results between frames.

[0203] Step 1.4, the discrimination of interactive vehicle motion cognition, mainly distinguishes whether the interactive vehicle is in a rational driving state, providing a priori judgment to facilitate the selection of subsequent decision-making models. Optionally, training data samples can be constructed: distinguishing different speed ranges (0-20km / h; 20-40km / h; 40-60km / h; above 60km / h), using the collision distance G between the front of the interactive vehicle and the merging endpoint as a classification feature, and collecting the G sequence of the interactive vehicle within a time domain.

[0204]

[0205] Where ΔD1 represents the relative longitudinal distance between the interactive vehicle and the merging endpoint; v1 represents the longitudinal speed of the interactive vehicle.

[0206] Then, the sample sequences are clustered using sequence clustering, and the clustering results represent rational driving states and irrational driving states, respectively.

[0207] Using the methods described above, a preliminary determination can be made. Figure 2 Whether vehicle B is in a rational driving state will help in selecting different vehicle interaction decision models in subsequent step 2.

[0208] Step 2, Vehicle Dynamic Interaction Decision-Making, mainly includes four parts: determining the right-of-way for the primary vehicle, vehicle interaction decision-making under rational driving conditions, vehicle interaction decision-making under irrational driving conditions, and parameter correction of the interaction decision-making model. Based on the results of Step 1.4, if vehicle B is in a rational driving state, the rational driving condition vehicle interaction decision-making model is selected; if vehicle B is in an irrational driving state, the irrational driving condition vehicle interaction decision-making model is selected.

[0209] Step 2.1, determining the right-of-way for the primary vehicle, i.e., determining the current vehicle's right-of-way. Since different right-of-way conditions affect the vehicle's decision-making process—for example, the interactive actions taken by the current vehicle may differ on roads with different right-of-way conditions—it is necessary to determine the right-of-way relationship between the current vehicle and interacting vehicles in advance. Optionally, the right-of-way relative to the interacting vehicles can be determined according to traffic regulations. Simultaneously, a lane-change decision is initiated for cases where the current vehicle has a low right-of-way. If the current vehicle initiates a lane-change decision and the lane change is successful, the process does not proceed to subsequent modules; if the lane change fails, the process proceeds to subsequent modules. Figure 2 It can be determined that vehicle A has high right-of-way and vehicle B has low right-of-way.

[0210] Step 2.2, Vehicle interaction decision-making under rational driving conditions: Based on the results of step 1.4, if vehicle B is in a rational driving state, then the vehicle interaction decision-making stage under rational driving conditions will begin.

[0211] Based on the theory of non-cooperative static games, an interactive model is constructed for vehicle A and vehicle B. First, the interaction action space of the vehicles is determined. The options are: longitudinal yielding and lateral avoidance, longitudinal yielding without lateral avoidance, longitudinal not yielding and lateral avoidance, and longitudinal not yielding and lateral not avoiding. Next, based on vehicle safety and efficiency, the cost functions C for vehicle A and vehicle B in the interaction model are designed respectively. a and C b Based on the current vehicle right-of-way status in step 2.1, determine the cost function C. a The weighting factor λ a Based on the cognitive results of vehicle B in step 1.4, we initially hypothesize that C b The prior value of the weighting factor is λ. b Finally, using the equilibrium solution method of the game theory model, we obtain the choices made by vehicle A and vehicle B. The probability values ​​of different interactive actions are used, and the interactive action corresponding to the highest probability value is taken as the output, representing the longitudinal and lateral decision-making results of the current vehicle and the longitudinal and lateral decision-making results of the interacting vehicle, respectively.

[0212] Step 2.3, Vehicle interaction decision-making in irrational driving situations: Based on the results of step 1.4, if vehicle B is in an irrational driving state, then the vehicle interaction decision-making stage in irrational driving situations will begin.

[0213] Here, we presuppose two types of irrational driving behavior in interactive vehicles: aggressive driving behavior and conservative driving behavior. First, based on the analysis of the current vehicle's right-of-way in step 2.1 and the judgment of the degree of interaction between vehicles in step 1.3, if the current vehicle is in a high right-of-way position and is significantly closer to the merging point than the interacting vehicle, then the current vehicle will choose to not yield longitudinally and not avoid laterally; if the current vehicle is in a low right-of-way position and is significantly farther from the merging point than the interacting vehicle, then the current vehicle will choose to yield longitudinally and avoid laterally; the remaining cases will be analyzed according to the irrational situation of the interacting vehicle: (1) If the driving behavior of the interacting vehicle is aggressive driving behavior, then it is believed that the interacting vehicle will take a longitudinal non-yield decision, so that the current vehicle can be controlled to choose a longitudinal yield decision. The lateral decision result is determined based on the relative lateral distance between the current vehicle and the interacting vehicle, that is, whether to avoid laterally can be selected based on the relative lateral distance between the current vehicle and the interacting vehicle; (2) If the driving behavior of the interacting vehicle is conservative driving behavior, then it is believed that the interacting vehicle will take a longitudinal yield decision, so that the current vehicle can be controlled to choose a longitudinal non-yield decision. The lateral decision result can also be determined based on the relative lateral distance between the current vehicle and the interacting vehicle, that is, whether to avoid laterally can be selected based on the relative lateral distance between the current vehicle and the interacting vehicle.

[0214] Step 2.4, parameter correction of the vehicle interaction decision model. Due to the uncertainty of human drivers, the actual driving behavior of the interactive vehicle will differ from the driving behavior preset by the decision model. Therefore, it is necessary to adjust the parameters of the decision model so that the prediction results of the interactive vehicle output by the decision model are consistent with its actual driving results.

[0215] This mainly adopts a feedback correction mechanism of "prior hypothesis - introduction of observation - posterior correction", and the main process is as follows: In step 2.2, it is known that C b The prior value of the weighting factor is λ. b After the decision-making model starts running, it determines whether the actual interactive action of vehicle B in the previous frame belongs to the interaction action space. In which of the following scenarios, i.e., when actual observation data is introduced, if the actual interactive actions of vehicle B are consistent with the longitudinal and lateral decision results of the interactive vehicle output by the decision model in the previous frame in step 2.2, then no correction is made; otherwise, λ is adjusted according to the actual interactive actions of vehicle B. bThis makes the new λ b The longitudinal and lateral decision-making results of the interactive vehicle output by the decision-making model are consistent with the actual interactive actions of vehicle B.

[0216] Step 3, Interactive Vehicle Trajectory Prediction. Based on the longitudinal and lateral decision results of the current vehicle and the interactive vehicle prediction results output in Step 2.2, the trajectory of the interactive vehicle is predicted to facilitate subsequent path and speed planning for the current vehicle. The trajectory prediction method adopted can be trajectory prediction based on the vehicle kinematics model, so that the predicted trajectory can conform to the interactive vehicle prediction results in Step 2.

[0217] Subsequently, the longitudinal and lateral decision results of vehicle A output in step 2 and the trajectory prediction results of vehicle B output in step 3 can be input into the vehicle path planning module to output the vehicle path planning result. The vehicle speed planning module then performs speed planning for the current vehicle based on the longitudinal and lateral decision results of vehicle A output in step 2, the trajectory prediction results of vehicle B output in step 3, and the vehicle path planning result, thereby obtaining the future trajectory planning point sequence of the current vehicle, which is then sent to the underlying control module for execution.

[0218] The following is combined with Figure 11 The longitudinal and lateral decision-making devices of the vehicle provided in this disclosure will be explained.

[0219] Figure 11 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the eleventh embodiment of this disclosure.

[0220] like Figure 11 As shown, the vehicle's longitudinal and lateral decision-making device 1100 may include: a first determining module 11, a first acquiring module 12, a second acquiring module 13, and a generating module 14.

[0221] The first determining module 11 is used to determine the dynamic interaction scenario of the current vehicle and to determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario.

[0222] The first acquisition module 12 is used to acquire the driving status of the interactive vehicle;

[0223] The second acquisition module 13 is used to acquire the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle.

[0224] The generation module 14 is used to generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle.

[0225] In one possible implementation of this disclosure, the first determining module 11 is specifically used for:

[0226] Acquire road images captured by the current vehicle;

[0227] The road features are obtained from the road images, and the dynamic interaction scenario of the current vehicle is determined based on the road features.

[0228] In one possible implementation of this disclosure, the first determining module 11 is specifically used for:

[0229] Acquire other vehicles in the dynamic interactive scene;

[0230] Calculate the distance between the current vehicle and other vehicles in the dynamic interaction scenario;

[0231] The vehicle that is closest to the current vehicle among the other vehicles is selected as the interaction vehicle.

[0232] In one possible implementation of this disclosure, the first acquisition module 12 is specifically used for:

[0233] Obtain the destination of the road where the interactive vehicle is located;

[0234] Obtain the collision distance G between the front of the interacting vehicle and the vehicle merging into the endpoint;

[0235] Obtain the G sequence of the interactive vehicles within a preset time period, and cluster the G sequence to generate clustering results;

[0236] The driving status of the interactive vehicle is obtained based on the clustering results.

[0237] In one possible implementation of this disclosure, the driving state of the interactive vehicle includes a rational driving state and an irrational driving state.

[0238] Based on the previous embodiment, this disclosure also provides a possible implementation of a vehicle's longitudinal and lateral decision-making device. Figure 12 The diagram below shows the structure of the longitudinal and lateral decision-making device for a vehicle according to the twelfth embodiment of this disclosure. Based on the previous embodiment, the longitudinal and lateral decision-making device 1200 for a vehicle further includes: a third acquisition module 15, a first processing module 16, a second processing module 17, and a third processing module 18.

[0239] The third acquisition module 15 is used to acquire the head-on collision time distance between the current vehicle and the interacting vehicle.

[0240] The first processing module 16 is used to further obtain the vehicle interaction decision model corresponding to the interactive vehicle based on the driving state of the interactive vehicle if the head-on collision distance is less than the first threshold.

[0241] The second processing module 17 is used to stop the decision if the distance to the frontal collision is greater than a second threshold, wherein the second threshold is greater than the first threshold.

[0242] The third processing module 18 is used to take the longitudinal and lateral decision results of the previous frame as the longitudinal and lateral decision results of the vehicle in the current frame if the collision distance at the front of the vehicle is greater than or equal to the first threshold and less than or equal to the second threshold.

[0243] Based on the above embodiments, this disclosure also provides a possible implementation of a vehicle's longitudinal and lateral decision-making device. Figure 13 This is a schematic diagram of the longitudinal and lateral decision-making device for a vehicle provided according to the thirteenth embodiment of this disclosure. Figure 11 Based on the embodiment, the vehicle's longitudinal and lateral decision-making device 1300 further includes: a fourth acquisition module 19 and a second determination module 20.

[0244] The fourth acquisition module 19 is used to acquire the road to which the current vehicle belongs, and the road to which the interactive vehicle belongs;

[0245] The second determining module 20 is used to determine the right-of-way relationship between the current vehicle and the interacting vehicle based on the road to which the current vehicle belongs and the road to which the interacting vehicle belongs.

[0246] In one possible implementation of this disclosure, if the driving state of the interactive vehicle is a rational driving state, then the generation module 14 is specifically used for:

[0247] Obtain the interaction space between the current vehicle and the interacting vehicles;

[0248] Generate the first cost function for the current vehicle and the second cost function for the interacting vehicle, respectively;

[0249] The first weight factor of the first cost function is adjusted according to the right-of-way relationship, and the second weight factor of the second cost function is set to the prior value.

[0250] The longitudinal and lateral decision results of the current vehicle are generated based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor.

[0251] In one possible implementation of this disclosure, if the driving state of the interactive vehicle is irrational, then the generation module 14 is specifically used for:

[0252] Obtain the first distance of the current vehicle relative to the merging destination, and the second distance of the interacting vehicle relative to the merging destination;

[0253] If the current vehicle is determined to have high right-of-way based on right-of-way relationships, and the first distance is greater than the second distance, then the current vehicle is controlled to choose the strategy of not yielding longitudinally and not avoiding laterally.

[0254] If the current vehicle is determined to have low right-of-way based on right-of-way relationships, and the first distance is less than or equal to the second distance, then the current vehicle is controlled to select a longitudinal yielding and lateral avoidance strategy.

[0255] In one possible implementation of this disclosure, if the driving state of the interactive vehicle is irrational, then the generation module 14 is specifically used for:

[0256] Acquire the driving behavior of the interactive vehicle;

[0257] If the driving behavior is aggressive, the current vehicle is controlled to select a longitudinal yielding strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle.

[0258] If the driving behavior is conservative, the current vehicle is controlled to select a longitudinal non-yield strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle.

[0259] Based on the previous embodiment, this disclosure also provides a possible implementation of a vehicle's longitudinal and lateral decision-making device. Figure 14 The diagram below shows the structure of the longitudinal and lateral decision-making device for a vehicle according to the fourteenth embodiment of this disclosure. Based on the previous embodiment, the longitudinal and lateral decision-making device 1400 for the vehicle further includes a fifth acquisition module 21 and an adjustment module 22.

[0260] Among them, the fifth acquisition module 21 is used to acquire the actual interactive actions of the interactive vehicle;

[0261] The adjustment module 22 is used to adjust the second weighting factor based on the actual interactive actions of the interactive vehicle and the predicted interactive actions of the interactive vehicle.

[0262] The longitudinal and lateral decision-making device for vehicles provided in this disclosure determines the current vehicle's dynamic interaction scenario and the interactive vehicles interacting with the current vehicle within that scenario. This allows the acquisition of the interactive vehicles' driving states, the generation of corresponding vehicle interaction decision models based on these driving states, and the generation of longitudinal and lateral decision-making results for the current vehicle based on these models. This addresses the problem of fragmented longitudinal and lateral decision-making in related technologies, while enabling the current vehicle to handle various uncertainties in complex dynamic interaction scenarios, such as the dynamic uncertainty of human driving behavior and the dynamic uncertainty of inter-vehicle interactions, thus ensuring the driving safety of the current vehicle in relation to the interactive vehicles.

[0263] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, a computer program product, and an autonomous vehicle.

[0264] Figure 15 A schematic block diagram of an example electronic device 1500 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0265] like Figure 15 As shown, the electronic device 1500 includes a computing unit 1501, which can perform various appropriate actions and processes according to a computer program stored in ROM (Read-Only Memory) 1502 or loaded from storage unit 1508 into RAM (Random Access Memory) 1503. The RAM 1503 may also store various programs and data required for the operation of the device 1500. The computing unit 1501, ROM 1502, and RAM 1503 are interconnected via bus 1504. An I / O (Input / Output) interface 1505 is also connected to bus 1504.

[0266] Multiple components in electronic device 1500 are connected to I / O interface 1505, including: input unit 1506, such as keyboard, mouse, etc.; output unit 1507, such as various types of displays, speakers, etc.; storage unit 1508, such as disk, optical disk, etc.; and communication unit 1509, such as network card, modem, wireless transceiver, etc. Communication unit 1509 allows device 1500 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0267] The computing unit 1501 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1501 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 1501 performs the various methods and processes described above, such as the aforementioned vehicle longitudinal and lateral decision-making method. For example, in some embodiments, the aforementioned vehicle longitudinal and lateral decision-making method can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 1508. In some embodiments, part or all of the computer program can be loaded and / or installed on device 1500 via ROM 1502 and / or communication unit 1509. When the computer program is loaded into RAM 1503 and executed by computing unit 1501, one or more steps of the vehicle longitudinal and lateral decision-making method described above can be performed. Alternatively, in other embodiments, computing unit 1501 can be configured to perform the vehicle longitudinal and lateral decision-making method by any other suitable means (e.g., by means of firmware).

[0268] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0269] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0270] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0271] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0272] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0273] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service ecosystem, addressing the shortcomings of traditional physical hosts and VPS (Virtual Private Server, or simply "VPS") services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0274] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0275] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0276] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for making longitudinal and lateral decision-making for vehicles, characterized in that, include: Determine the dynamic interaction scenario of the current vehicle, and determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario. The driving state of the interactive vehicle is obtained; the driving state of the interactive vehicle includes rational driving state and irrational driving state. The vehicle interaction decision model corresponding to the interactive vehicle is obtained based on the driving status of the interactive vehicle. The longitudinal and lateral decision results of the current vehicle are generated based on the vehicle interaction decision model corresponding to the interactive vehicle; This also includes: Obtain the road to which the current vehicle belongs, and the road to which the interacting vehicle belongs; The right-of-way relationship between the current vehicle and the interacting vehicle is determined based on the road to which the current vehicle belongs and the road to which the interacting vehicle belongs; Wherein, if the driving state of the interactive vehicle is irrational driving, then generating the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle includes: Obtain a first distance of the current vehicle relative to the merging endpoint of the road where the interactive vehicle is located, and a second distance of the interactive vehicle relative to the merging endpoint; If the current vehicle is determined to have high right-of-way based on the right-of-way relationship, and the first distance is greater than the second distance, then the current vehicle is controlled to select a longitudinal non-yield and lateral non-avoidance strategy. If the current vehicle is determined to have low right-of-way based on the right-of-way relationship, and the first distance is less than or equal to the second distance, then the current vehicle is controlled to select a longitudinal yielding and lateral avoidance strategy.

2. The method as described in claim 1, characterized in that, Determining the current dynamic interaction scenario of the vehicle includes: Acquire the road image captured by the current vehicle; The road features are obtained from the road image, and the dynamic interaction scenario of the current vehicle is determined based on the road features.

3. The method as described in claim 1, characterized in that, The interactive vehicles that interact with the current vehicle in the dynamic interactive scenario include: Acquire other vehicles in the dynamic interactive scene; Calculate the distance between the current vehicle and other vehicles in the dynamic interaction scenario; The vehicle that is closest to the current vehicle among the other vehicles is designated as the interactive vehicle.

4. The method as described in claim 1, characterized in that, After determining the interactive vehicle that interacts with the current vehicle in the dynamic interactive scenario, the method further includes: Obtain the head-on collision time distance between the current vehicle and the interacting vehicle; If the distance to the frontal collision is less than the first threshold, then the vehicle interaction decision model corresponding to the interactive vehicle is further obtained based on the driving state of the interactive vehicle. If the distance to the frontal collision is greater than the second threshold, the decision is stopped, wherein the second threshold is greater than the first threshold; If the time distance of the frontal collision is greater than or equal to the first threshold and less than or equal to the second threshold, then the longitudinal and lateral decision results of the previous frame are used as the longitudinal and lateral decision results of the current vehicle in the current frame.

5. The method as described in claim 1, characterized in that, The step of obtaining the driving status of the interactive vehicle includes: Obtain the merging endpoint of the road where the interactive vehicle is located; Obtain the head-on collision distance G of the interactive vehicle relative to the merging endpoint; Obtain the G sequence of the interactive vehicle within a preset time period, and cluster the G sequence to generate clustering results; The driving status of the interactive vehicle is obtained based on the clustering results.

6. The method as described in claim 1, characterized in that, If the driving state of the interactive vehicle is rational driving, then generating the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle includes: Obtain the interaction space between the current vehicle and the interacting vehicle; Generate the first cost function of the current vehicle and the second cost function of the interactive vehicle respectively; The first weight factor of the first cost function is adjusted according to the right-of-way relationship, and the second weight factor of the second cost function is set as a priori value. The longitudinal and lateral decision results of the current vehicle are generated based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor.

7. The method as described in claim 1, characterized in that, If the driving state of the interactive vehicle is irrational, then generating the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle includes: Obtain the driving behavior of the interactive vehicle; If the driving behavior is aggressive driving behavior, then the current vehicle is controlled to select a longitudinal yielding strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle. If the driving behavior is conservative, then the current vehicle is controlled to select a longitudinal non-yield strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle.

8. The method as described in claim 6, characterized in that, Also includes: Obtain the actual interactive actions of the interactive vehicle; The second weighting factor is adjusted based on the actual interactive actions of the interactive vehicle and the predicted interactive actions of the interactive vehicle.

9. A vehicle longitudinal and lateral decision-making device, characterized in that, include: The first determining module is used to determine the dynamic interaction scenario of the current vehicle and to determine the interactive vehicle that interacts with the current vehicle in the dynamic interaction scenario. The first acquisition module is used to acquire the driving state of the interactive vehicle; the driving state of the interactive vehicle includes rational driving state and irrational driving state. The second acquisition module is used to acquire the vehicle interaction decision model corresponding to the interactive vehicle based on the driving status of the interactive vehicle. The generation module is used to generate the longitudinal and lateral decision results of the current vehicle based on the vehicle interaction decision model corresponding to the interactive vehicle. This also includes: The fourth acquisition module is used to acquire the road to which the current vehicle belongs, and the road to which the interactive vehicle belongs; The second determining module is used to determine the right-of-way relationship between the current vehicle and the interactive vehicle based on the road to which the current vehicle belongs and the road to which the interactive vehicle belongs; Wherein, if the driving state of the interactive vehicle is irrational driving, the generation module is specifically used for: Obtain a first distance of the current vehicle relative to the merging endpoint of the road where the interactive vehicle is located, and a second distance of the interactive vehicle relative to the merging endpoint; If the current vehicle is determined to have high right-of-way based on the right-of-way relationship, and the first distance is greater than the second distance, then the current vehicle is controlled to select a longitudinal non-yield and lateral non-avoidance strategy. If the current vehicle is determined to have low right-of-way based on the right-of-way relationship, and the first distance is less than or equal to the second distance, then the current vehicle is controlled to select a longitudinal yielding and lateral avoidance strategy.

10. The apparatus as claimed in claim 9, characterized in that, The first determining module is specifically used for: Acquire the road image captured by the current vehicle; The road features are obtained from the road image, and the dynamic interaction scenario of the current vehicle is determined based on the road features.

11. The apparatus as claimed in claim 9, characterized in that, The first determining module is specifically used for: Acquire other vehicles in the dynamic interactive scene; Calculate the distance between the current vehicle and other vehicles in the dynamic interaction scenario; The vehicle that is closest to the current vehicle among the other vehicles is designated as the interactive vehicle.

12. The apparatus as claimed in claim 9, characterized in that, Also includes: The third acquisition module is used to acquire the head-on collision time distance between the current vehicle and the interacting vehicle; The first processing module is used to further obtain the vehicle interaction decision model corresponding to the interactive vehicle based on the driving state of the interactive vehicle if the collision distance at the front of the vehicle is less than a first threshold. The second processing module is used to stop the decision if the distance to the frontal collision is greater than a second threshold, wherein the second threshold is greater than the first threshold. The third processing module is used to take the longitudinal and lateral decision results of the previous frame as the longitudinal and lateral decision results of the current vehicle in the current frame if the collision distance is greater than or equal to the first threshold and less than or equal to the second threshold.

13. The apparatus as claimed in claim 9, characterized in that, The first acquisition module is specifically used for: Obtain the merging endpoint of the road where the interactive vehicle is located; Obtain the head-on collision distance G of the interactive vehicle relative to the merging endpoint; Obtain the G sequence of the interactive vehicle within a preset time period, and cluster the G sequence to generate clustering results; The driving status of the interactive vehicle is obtained based on the clustering results.

14. The apparatus as claimed in claim 9, characterized in that, If the driving state of the interactive vehicle is a rational driving state, then the generation module is specifically used for: Obtain the interaction space between the current vehicle and the interacting vehicle; Generate the first cost function of the current vehicle and the second cost function of the interactive vehicle respectively; The first weight factor of the first cost function is adjusted according to the right-of-way relationship, and the second weight factor of the second cost function is set as a priori value. The longitudinal and lateral decision results of the current vehicle are generated based on the first cost function, the second cost function, the first weighting factor, and the second weighting factor.

15. The apparatus as claimed in claim 9, characterized in that, If the driving state of the interactive vehicle is irrational, then the generation module is specifically used for: Obtain the driving behavior of the interactive vehicle; If the driving behavior is aggressive driving behavior, then the current vehicle is controlled to select a longitudinal yielding strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle. If the driving behavior is conservative, then the current vehicle is controlled to select a longitudinal non-yield strategy, and a lateral avoidance strategy is selected based on the relative lateral distance between the current vehicle and the interacting vehicle.

16. The apparatus as claimed in claim 14, characterized in that, Also includes: The fifth acquisition module is used to acquire the actual interactive actions of the interactive vehicle; The adjustment module is used to adjust the second weighting factor based on the actual interactive actions of the interactive vehicle and the predicted interactive actions predicted for the interactive vehicle.

17. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-8.

18. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-8.

19. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-8.

20. An autonomous vehicle, including the electronic equipment as claimed in claim 17.

Images