Intelligent driving method and vehicle to which the method is applied

Abstract

The present application discloses an intelligent driving method and a vehicle to which the method is applied. A judgment solution set of the host vehicle may be obtained based on the recognition information, and a first judgment solution is obtained from the judgment solution set for a first interaction stage between the host vehicle and the conflicting object, and in the process of controlling the host vehicle to run based on the first judgment solution, it is detected that the host vehicle and the conflicting object meet the condition to enter a second interaction stage, and a second judgment solution is obtained from the judgment solution set for a second interaction stage between the host vehicle and the conflicting object, and the host vehicle is controlled to run based on the second judgment solution. In this way, dynamic vehicle judgment is implemented, the problem that the judgment result is inconsistent with the actual behavior of the conflicting object is effectively handled, the generalization ability of the conflict judgment is improved, and the safety of intelligent driving and the comfort of driving and riding are improved.

Description

[Technical field]

[0001] The present application relates to the field of vehicle technology, and in particular to an intelligent driving method and a vehicle to which the method is applied. [Background technology]

[0002] In the field of autonomous driving, there are many cases where the self-vehicle interacts with another vehicle during the driving process. At present, in the scene where the self-vehicle interacts with another vehicle, the mainstream judgment method is that the self-vehicle obtains a judgment solution based on the predicted path of the other vehicle, and drives according to the judgment solution.

[0003] In a practical case, as shown in FIG. 1(a), assume that another vehicle is waiting to turn left and the host vehicle is going straight. According to the conventional rule of going straight and yielding to the passing vehicle before turning left, the host vehicle predicts that the other vehicle will yield to the host vehicle. Based on the predicted path of the other vehicle acquired by the host vehicle, the obtained judgment solution may be that the host vehicle accelerates and passes quickly. However, in the actual interaction process, as shown in FIG. 1(b), the future behavior of the other vehicle is not necessarily that of yielding based on the predicted path. If the host vehicle still drives according to the previously acquired judgment solution, the probability of the host vehicle colliding or scraping with the other vehicle is relatively high, and the probability of the host vehicle passing safely is relatively low. In addition to the above, when the host vehicle will collide with the other vehicle, the host vehicle or the driver may brake urgently (for example, the host vehicle activates the AEB (autonomous emergency braking) function to avoid the collision), which causes a poor driving feeling. Furthermore, the prediction results of the vehicle may not match the actual situation, so different results may be obtained even in the same scene, which makes the generalization of the autonomous driving control policy relatively poor.

[0004] In summary, methods are required to improve interaction safety, driving and riding comfort, and generalization of control policies for autonomous driving. Summary of the Invention [Means for solving the problem]

[0005] In the present application, an intelligent driving method and a vehicle to which the method is applied are provided, which improves interaction safety, driving and riding comfort, and generalization of control policies for vehicle autonomous driving.

[0006] The intelligent driving methods provided in this application may be performed by an electronic device, which may be abstracted as a computer system. The electronic device may be an entire system or a component part of an entire system, such as a system chip or a processing chip.

[0007] In particular, the electronic device may be a terminal device or an in-vehicle device such as an in-vehicle computer or in-vehicle infotainment in a vehicle, or may be a system chip, a decision processing chip, or another type of chip that can be disposed in a computer system of the vehicle or in-vehicle device.

[0008] According to a first aspect, an embodiment of the present application provides a method for producing a method for detecting a pulmonary circulation of a lung cancer cell comprising: a step of acquiring a judgment solution set for the host vehicle based on the recognition information; a step of acquiring a first judgment solution from the judgment solution set to be used in a first interaction stage between the host vehicle and the competing object; a step of detecting that the host vehicle and the competing object satisfy a condition for entering a second interaction stage in a process of controlling the host vehicle to travel based on the first judgment solution; a step of acquiring a second judgment solution from the judgment solution set to be used in a second interaction stage between the host vehicle and the competing object; and a step of controlling the host vehicle to travel based on the second judgment solution. An intelligent driving method is provided, including:

[0009] The recognition information may be obtained based on an in-vehicle recognition device of the vehicle. The in-vehicle recognition device may include, but is not limited to, a LIDAR, a millimeter wave radar, an ultrasonic radar, or a camera. The recognition information may also be obtained by a method such as V2X (Vehicle to Everything), V2V (Vehicle to Vehicle), or V2I (Vehicle to Infrastructure).

[0010] The first interaction stage is a stage in which the interaction relationship between the host vehicle and the competing object is unclear, and the second interaction stage is a stage in which the interaction relationship between the host vehicle and the competing object is clear.

[0011] For example, when the above method is performed by an electronic device, the recognition information may be acquired using a collection device of the electronic device, and the collection device may send the recognition information to the processing device of the electronic device using an interface circuit connected to the processing device, or the recognition information may be acquired using a collection device externally connected to the electronic device, and the collection device may send the recognition information to the processing device of the electronic device via a wired or wireless connection manner. The processing device may acquire a judgment solution set for the host vehicle based on the recognition information, acquire a first judgment solution from the judgment solution set for use in a first interaction stage between the host vehicle and the competing object, and detect that the host vehicle and the competing object satisfy a condition for entering a second interaction stage, acquire a second judgment solution from the judgment solution set for use in a second interaction stage between the host vehicle and the competing object, and control the host vehicle to run based on the second judgment solution in the process of controlling the host vehicle to run based on the first judgment solution.

[0012] According to the method of the present application, in the vehicle judgment process, the judgment process is divided into multiple interaction stages, the competing object is continuously tracked and recognized, and the vehicle interaction stage is continuously updated, so that dynamic vehicle judgment is implemented, the problem that the judgment result is not consistent with the actual behavior of the competing object can be effectively handled, the automatic driving turn rate is reduced, and the generalization ability of the conflict judgment is improved. In addition, since the judgment process includes multiple interaction stages, the vehicle advances step by step throughout the judgment process, so that the sudden acceleration and deceleration are effectively reduced, and the comfort of the automatic driving is improved.

[0013] In a possible design, the first interaction stage of the present application can be understood as a crisis state. The crisis state can be understood as an interaction relationship between the self-vehicle and the competing object being unclear. The interaction relationship being unclear can be understood as the benefits of multiple judgment solutions of the interaction between the self-vehicle and the competing object not being significantly different, that is, the judgment solution of the self-vehicle may change at any time, that is, it may jump from judgment solution A to judgment solution B.

[0014] When the vehicle and the conflicting object are in a crisis state, the vehicle uses a first judgment solution to continuously track the status of the conflicting object in the traveling process (which can be understood as viewing it one step at a time), and the judgment solution may be adjusted at any time.

[0015] In a possible design, the second interaction stage of the present application can be understood as a trust state. The trust state can be understood as an interaction relationship between the host vehicle and the competing object being clear. The interaction relationship being clear can be understood as a judgment solution of the interaction between the host vehicle and the competing object being determined.

[0016] For example, in a competition process between a vehicle and a competing target, the vehicle approaches a superior judgment solution as the competition process continues. When the difference between the superior judgment solution and another judgment solution is large enough (usually, the difference can be determined using the profit difference between different judgment solutions), the competitive relationship is considered clear, and the transition from the crisis state to the trust state is completed.

[0017] In a possible design, during the process in which the ego-vehicle performs the first interaction phase with the conflicting object, the status of the conflicting object is continuously tracked.

[0018] In a possible design, the conditions for entering the second interaction phase include one of: the duration of the first interaction phase exceeds a threshold duration; the decision benefit difference between the decision solution with the first largest decision benefit and the decision solution with the second largest decision benefit in the decision solution set is not less than a threshold benefit difference; the actual movement path of the competing object at a given time point is not consistent with the first decision solution; the actual movement path of the competing object at a given time point is consistent with the first decision solution; or the distance between the ego vehicle and the competing object is less than a threshold distance.

[0019] According to the method, the vehicle judgment process provides multiple conditions for deciding to enter the second interaction stage, which is more flexible and more adaptable.

[0020] In a possible design, at least one object is determined based on the recognition information, a future movement path of the at least one object is obtained, and among the at least one object, an object whose future movement path intersects with the future movement path of the vehicle is determined as a competing object.

[0021] For example, when the above method is performed by an electronic device, the recognition information may be acquired using a collection device of the electronic device, and the collection device may send the recognition information to the processing device of the electronic device using an interface circuit connected to the processing device, or the recognition information may be acquired using a collection device externally connected to the electronic device, and the collection device may send the recognition information to the processing device of the electronic device via a wired or wireless connection. The processing device may determine at least one object based on the recognition information, and may determine competing objects from the at least one object.

[0022] By using this method, a scheme for determining competing targets in the vehicle decision process is provided, which is highly adaptive.

[0023] In a possible design, a left / right position offset sampling range and a front / rear acceleration sampling range of the ego-vehicle and the competing object are obtained based on the recognition information, and a judgment solution set is obtained based on the left / right position offset sampling range and the front / rear acceleration sampling range.

[0024] For example, when the above-mentioned method is performed by an electronic device, the processing device of the electronic device may obtain a left-right position offset sampling range and a front-back acceleration sampling range of the host vehicle and the competing object based on the recognition information, and then the processing device of the electronic device obtains a judgment solution set based on the left-right position offset sampling range and the front-back acceleration sampling range.

[0025] In a possible design, a left / right position offset sampling range of the host vehicle and the competing object is obtained based on the recognition information and a first factor, and a forward / backward acceleration sampling range of the host vehicle and the competing object is obtained based on the recognition information and a second factor, the first factor including at least one of road boundaries and stationary obstacles of the environment in which the host vehicle is located, and the second factor including at least one of road speed limits, jerk, and competing object type of the environment in which the host vehicle is located.

[0026] For example, when the above-mentioned method is performed by an electronic device, a processing unit of the electronic device may obtain a left-right position offset sampling range of the host vehicle and the competing object based on the recognition information and a first factor, and a processing unit of the electronic device may obtain a forward-backward acceleration sampling range of the host vehicle and the competing object based on the recognition information and a second factor.

[0027] According to the present invention, a method is provided for determining a lateral position offset sampling range and a longitudinal acceleration sampling range in a vehicle determination process.

[0028] In a possible design, at least one judgment solution subset is obtained by dividing the judgment solution set, each judgment solution subset includes at least one judgment solution, an optimal judgment solution corresponding to each judgment solution subset is determined, the optimal judgment solution is a judgment solution with a maximum judgment profit in the corresponding judgment solution subset, and from the at least one optimal judgment solution, a first judgment solution to be used in a first interaction stage between the ego-vehicle and the competing target is obtained.

[0029] According to the present invention, a method for selecting a first judgment solution in a vehicle judgment process is provided, and a first judgment solution is selected from optimal judgment solutions corresponding to different judgment solution subsets, so that the selected first judgment solution can be more practical, and interaction safety and traffic rate can be improved.

[0030] In a possible design, target judgment solutions whose judgment profit is not less than the threshold profit of at least one optimal judgment solution are classified into a first tier, and optimal judgment solutions whose judgment profit is less than the threshold profit are classified into a second tier. In other words, a judgment solution subset whose judgment profit exceeds the threshold profit (each judgment solution subset is "represented" by an optimal judgment solution corresponding to the judgment solution subset) is determined as the first tier, and a judgment solution subset whose judgment profit is less than the threshold profit is determined as the second tier. From the optimal judgment solutions included in the first tier, a first judgment solution to be used for a first interaction stage between the ego vehicle and the competing target is determined.

[0031] The method further provides a method for selecting a first judgment solution in a vehicle judgment process. The judgment solutions are more precisely divided based on a profit threshold to obtain a first hierarchy with a relatively high judgment profit, and then the first judgment solution is determined from the optimal judgment solutions included in the first hierarchy. This more effectively improves the applicability of the first judgment solution, and further improves the interaction safety and the passage rate. In a possible design, a second judgment solution is obtained from the target judgment solutions included in the first hierarchy, which is used for a second interaction stage between the ego-vehicle and the competing target.

[0032] In the possible design, a judgment benefit corresponding to each judgment solution in the judgment solution subset is determined based on a benefit evaluation scheme, and a judgment solution with a maximum judgment benefit in the judgment solution subset is determined as a target judgment solution corresponding to the judgment solution subset. The benefit evaluation scheme includes at least one magnitude of safety, comfort, passability, right-of-way, danger area, and offset of the judgment solution.

[0033] For example, when the above method is performed by an electronic device, a processing unit of the electronic device may determine a judgment benefit corresponding to each judgment solution in the judgment solution subset according to a benefit evaluation scheme.

[0034] Using this method, a method for determining the decision benefit of a decision solution is provided in the vehicle decision process.

[0035] In the possible design, based on the correspondence between the driving environment and the benefit evaluation method, the benefit evaluation method used for the interaction is determined, and the judgment benefit corresponding to each judgment solution in the judgment solution subset is determined according to the benefit evaluation method used for the interaction.

[0036] According to this method, the present application provides a correspondence between the driving environment and the benefit evaluation method, so that a more suitable benefit evaluation method can be selected in the vehicle judgment process by referring to the current driving environment of the vehicle, so that the obtained judgment benefit of the judgment solution is more realistic, more targeted, more authentic and effective.

[0037] In a possible design, before a first judgment solution to be used in a first interaction stage with the competing target is determined from the judgment solution set, it is determined that the ego vehicle and the competing target satisfy conditions for entering the first interaction stage, the conditions for entering the first interaction stage including the difference in judgment benefits of any two target judgment solutions in the first tier being less than a threshold benefit difference.

[0038] According to the method, the conditions for entering the first interaction stage are provided in a vehicle decision process.

[0039] In a possible design, the method further includes determining to perform the second stage if a difference between the judgment benefit of the target judgment solution in the first tier and the judgment benefit of each remaining target judgment solution in the first tier exceeds a threshold benefit difference.

[0040] In a possible design, each decision solution includes one action pair group, and the action pair group includes vehicle speed and vehicle offset direction.

[0041] In a possible design, the method further includes adjusting the decision solution used in the first interaction phase if it is determined in the first interaction phase that a tracking result of the competing object at a predetermined time point is inconsistent with a movement path at a predetermined time point predicted using the first decision solution.

[0042] It can be seen that the judgment solution used for the interaction is not clear in the first interaction stage. Therefore, in the process in which the self-vehicle runs using the first judgment solution, the self-vehicle may continuously track the status of the conflicting object and continuously adjust the judgment solution.

[0043] In a possible design, the first decision solution is the same as the second decision solution.

[0044] As an example, the host vehicle uses the judgment solution a of the first interaction solution. Then, if it is determined that the judgment profit of the judgment solution a is the largest and the judgment profit difference between the judgment solution a and the judgment solution b with the second largest judgment profit exceeds the threshold profit difference, the host vehicle determines that the first interaction stage needs to be switched to the second interaction stage, and the host vehicle switches to the second interaction stage and drives using the judgment solution a.

[0045] According to a second aspect, the present application provides a computing device including a processor, connected to a memory, the memory storing computer programs or instructions, the processor configured to execute the computer programs or instructions stored in the memory, such that the computing device performs the method of any possible implementation of the first aspect.

[0046] According to a third aspect, the present application provides a computer readable storage medium storing a computer program or instructions which, when executed, enable a computer to perform the method of any possible implementation of the first aspect.

[0047] According to a fourth aspect, the present application provides a computer program product, which, when executed by a computer, enables the computer to perform the method of any possible implementation of the first aspect.

[0048] According to a fifth aspect, the present application provides a chip, connected to a memory and configured to read and execute computer programs or instructions stored in the memory to perform the method of any possible implementation of the first aspect.

[0049] According to a sixth aspect, the present application provides a vehicle, the vehicle including an interior perception device and an execution device, for performing the method of the first aspect or any one of the possible implementations of the first aspect.

[0050] According to a seventh aspect, the present application provides a vehicle, comprising the chip of the fifth aspect and an execution device for implementing the method of any possible implementation of the first aspect.

[0051] According to an eighth aspect, the present application provides a vehicle. The vehicle includes a recognition module configured to acquire recognition information. The vehicle further includes a processing module configured to acquire a judgment solution set of the host vehicle based on the recognition information, acquire a first judgment solution from the judgment solution set to be used in a first interaction stage between the host vehicle and a conflicting object, detect that the host vehicle and the conflicting object satisfy a condition for entering a second interaction stage in the process of controlling the host vehicle to run based on the first judgment solution, acquire a second judgment solution from the judgment solution set to be used in a second interaction stage between the host vehicle and the conflicting object, and control the host vehicle to run based on the second judgment solution.

[0052] It is to be understood that the technical solutions provided in this application may be applied to mass-produced products such as autonomous vehicles and mobile robots, thereby implementing dynamic judgment for interaction conflicts between the product and multiple obstacles (competing objects), providing a more applicable and reasonable judgment solution, effectively dealing with the problem that the judgment result is not consistent with the actual behavior of the competing objects, improving the safety of autonomous driving, and improving the generalization ability of the autonomous driving control policy, thereby better completing the interaction with multiple obstacles (competing objects) in complex scenes. In addition to the above, when the technical solutions are applied to autonomous driving scenes, the judgment process can include multiple interaction stages, so that the vehicle advances step by step through the judgment process, thereby effectively reducing sudden acceleration and deceleration, and improving the comfort of autonomous driving. [Brief description of the drawings]

[0053] [Figure 1] FIG. 2 is a schematic diagram of a vehicle judgment scene. [Diagram 2] 1 is a schematic diagram of the structure of an electronic device according to the present application; [Diagram 3] FIG. 2 is a schematic diagram of the composition of a processing module according to the present application. [Figure 4] FIG. 2 is a schematic diagram of the structure of a second electronic device according to the present application. [Diagram 5] FIG. 2 is a schematic diagram of the structure of a third electronic device according to the present application. [Figure 6] FIG. 2 is a schematic diagram of a conflict determination scene according to the present application; [Figure 7] FIG. 2 is a schematic diagram of another conflict determination scenario according to the present application; [Figure 8] FIG. 2 is a schematic diagram of generating left-right and front-back sampling policies according to the present application; [Figure 9] FIG. 2 is a schematic diagram of a decision benefit evaluation curve according to the present application. [Figure 10] FIG. 2 is a schematic diagram of a vehicle driving offset scene according to the present application; [Figure 11] 1 is a schematic diagram of a matching decision scene according to the present application; [Figure 12] 1 is a schematic flowchart of a first intelligent driving method according to the present application. [Figure 13] 4 is a schematic flowchart of a second intelligent driving method according to the present application. [Figure 14] FIG. 2 is a schematic diagram of a decision solution subset according to the present application. [Figure 15] FIG. 2 is a schematic diagram of a speed plan for the first interaction stage according to the present application. [Figure 16] FIG. 2 is a schematic diagram of a switching scene from a first interaction stage to a second interaction stage according to the present application. [Figure 17] FIG. 2 is a schematic diagram of another scenario for switching from a first interaction stage to a second interaction stage according to the present application; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0054] In the present application, an intelligent driving method and a vehicle to which the method is applied are provided. In particular, the intelligent driving method may be used in a vehicle decision-making process, so that the vehicle has a better ability to interact with surrounding obstacles in the autonomous driving process.

[0055] According to the method provided in the embodiment of the present application, in the vehicle judgment process, the difference between the actual movement path of a competing object (e.g., a competing vehicle) and the predicted movement path is taken into account, and the interaction process between the self-vehicle and the competing object is divided into multiple competition stages to implement dynamic vehicle judgment, so that the problem of the judgment result being inconsistent with the actual movement path of the competing object can be effectively handled, thereby reducing the automatic driving turnover rate and the possibility of accidents, improving the safety and comfort of automatic driving, and improving the generalization ability of competition judgment.

[0056] The electronic device of the embodiment of the present application can be used to support a vehicle in implementing the intelligent driving method provided in the embodiment of the present application.

[0057] Alternatively, the electronic device of the embodiment of the present application may be an in-vehicle device, and so on.

[0058] In an embodiment of the present application, the vehicle may communicate with another object based on a vehicle-to-external wireless communication technology (e.g., vehicle to everything (V2X)). For example, the communication between the vehicle and the other object may be performed based on a vehicle-to-vehicle wireless communication technology (e.g., vehicle to vehicle (V2V)). The communication between the vehicle and the other object may be performed based on Wi-Fi, 5th generation (5G) mobile communication technology, long term evolution (LTE), etc.

[0059] Classification standards for autonomous vehicles have been proposed by the industry. The driving automation classification standard proposed by the Society of Automotive Engineers International (SAE International) includes six levels, from L0 to L5. L0 to L2 indicate that the driver support system can provide some support functions to the driver. However, regardless of whether the driver support functions of the vehicle are disabled, the driver must drive the vehicle independently, constantly monitor the support functions provided by the driver support system, and perform steering, braking, and acceleration when required to ensure safety. The difference between L0 and L2 support functions is as follows: L0 indicates that no driving automation is provided, and the support functions are limited to warnings and short-term assistance; L1 support functions provide steering or braking / acceleration support to the driver; and L2 support functions provide steering and braking / acceleration support to the driver. L3 semi-autonomous driving: The autonomous driving system can complete some driving tasks and in some cases can monitor the driving environment. However, the driver must be ready to take driving control at any time. For example, the driver must drive as requested by the function. L4 Highly automated driving: The automated driving system can complete driving tasks and monitor the driving environment under specific conditions and in specific environments. L5 Fully automated driving: The automated driving system can complete all driving tasks under all conditions. The solutions of the embodiments of the present application can be applied to vehicles whose automated driving capabilities reach L2 or higher. Details will not be described below.

[0060] In addition to the above, it is understood that based on actual usage requirements, the technical solutions of the present application can also be applied to other carriers or transportation means, such as aircraft and mobile robots.

[0061] As an example, Fig. 2(a) is a schematic diagram of a possible electronic device architecture, which may include a processing module 210 and a communication module 220. Alternatively, the architecture shown in Fig. 2(a) may be an in-vehicle device or a functional component having electronic devices as described in the present application.

[0062] Alternatively, as shown in Fig. 2(b), the electronic device may further include a collection module 230 configured to collect recognition information. The collection module of the embodiment of the present application may include a device used for information collection function, such as a camera device or a radar device. The radar may include a LIDAR, a millimeter wave radar, an ultrasonic radar, etc.

[0063] It can be seen that the perception information obtained by the electronic device in the embodiment of the present application includes information about the host vehicle and perception information of the surrounding environment in which the host vehicle is located.

[0064] For example, the recognition information of an embodiment of the present application may include the position of the host vehicle, the current wheel steering of the host vehicle, the position of another vehicle within a threshold range of the host vehicle, the distance between the other vehicle and the host vehicle, the position of a road boundary within a threshold range of the host vehicle, the distance between the road boundary and the host vehicle, etc.

[0065] The threshold range of the present application may be the maximum recognition range of the collection device (e.g., a sensor) of the host vehicle. For example, if the recognition range of the collection device of the host vehicle is a circular area formed with a radius of 200 meters using the vehicle on which the collection device is located as the center, the threshold range of the present application may be a circular area formed with a radius of 200 meters using the host vehicle as the center.

[0066] Apart from the above, as shown in FIG. 2(c), the electronic device may further include an execution module 240 configured to execute the decision solution obtained by the processing module.

[0067] For example, when the vehicle is in an autonomous driving state, after obtaining the judgment solution determined by the processing module using the communication module of the electronic device, the execution module controls the vehicle to run according to the judgment solution.

[0068] If the electronic device does not include an execution module, the communication module of the electronic device may send the judgment solution to the execution module externally connected to the electronic device via a wired or wireless connection method (e.g., Bluetooth connection, network connection or interface circuit), thereby enabling the execution module to control the automatic driving of the vehicle according to the judgment solution.

[0069] Apart from the above, as shown in FIG. 2(d), the electronic device may further include a storage module 250 configured to store one or more programs and data information, where the one or more programs include instructions.

[0070] Apart from the above, as shown in FIG. 2(e), the electronic device may further include a display screen 260, which may display the judgment solution in real time.

[0071] If the structure is an in-vehicle device or another electronic device, the collection module 230 may include a device configured to support perception information collection functions, such as a camera device or a radar device. The processing module 210 may be a processor, such as a central processing unit (CPU). The collection module 230 may communicate with the processing module 210 through the communication module 220. The processing module 210 may perform the intelligent driving methods provided herein based on the acquired perception information.

[0072] If the structure is a functional component having electronic devices as described herein, the collection module 230 may include devices configured to support perception information collection functions, such as a camera device, a perception device, a radar device, etc. The processing module 210 may be a processor. The collection module 230 may communicate with the processing module 210 through the communication module 220. The processing module 210 may perform the intelligent driving method provided herein based on the acquired perception information.

[0073] When the structure is a chip or chip system, the collection module 230 may be one or more of a camera device, a recognition device, or a radar device controlled by the chip, and the processing module 210 may be a processor of the chip and may include one or more central processing units. It should be understood that the processing module 210 of this embodiment of the present application may be realized by a processor or a processor-related circuit component, and the collection module 230 may be realized by a related collection device such as a camera device, a recognition device, or a radar device.

[0074] The processing module 210 may be configured to perform all operations of the intelligent driving method performed by the electronic device of any embodiment of the present application, for example, to obtain a set of judgment solutions based on the acquired recognition information, and then determine a first judgment solution to be used for a first interaction stage with the conflicting object from the set of judgment solutions, continuously track the status of the conflicting object in the process of the first interaction stage with the conflicting object, and perform driving using a second judgment solution corresponding to the second interaction stage after determining that the host vehicle and the conflicting object meet the conditions for entering the second interaction stage.

[0075] The recognition information acquired by the processing module 210 may be one or more of image information or acoustic information collected by an external sensor or camera, and the recognition information acquired by the processing module 210 may be one or more of image information or acoustic information collected by a sensor or camera of the electronic device.

[0076] The processing module 210 may include multiple functional modules. For example, as shown in FIG. 3, the multiple functional modules included in the processing module may be a perception module, a judgment module, and a control module. The perception module is configured to perform analysis operations on the perception information and establish a world model including roads, obstacles, conflicting objects, etc. for downstream modules. The judgment module can generate judgment solutions based on the perception information to obtain a predicted path. The control module can perform autonomous driving based on the judgment solutions.

[0077] 4 is a schematic diagram of another electronic device structure configured to perform the operations performed by the electronic device according to an embodiment of the present application. As shown in FIG. 4, the electronic device may include a processor, a memory, and an interface circuit.

[0078] In addition to the above, the electronic device may further include another component, for example, a collection device. The processor is mainly configured to perform the processing operations provided in the embodiments of the present application, for example, configured to perform analysis processing on the acquired recognition information. The memory is mainly configured to store software programs and data. The collection device may be configured to collect the recognition information, etc. The interface circuit may be configured to support communication of the electronic device. For example, after the collection device collects the recognition information, the collection device may send the collected recognition information to the processor using the interface circuit. The interface circuit may include a transceiver or an input / output interface.

[0079] Further, as shown in Fig. 5, an embodiment of the present application further provides a vehicle judgment system. The vehicle judgment system may include a competing obstacle screening module, a competing object status tracking module, a competing judgment module, a motion planning module, etc.

[0080] The competing obstacle screening module is configured to obtain the perception information and determine a competing object based on the perception information.

[0081] Alternatively, the recognition information obtained by the competing obstacle screening module may be collected by a collection module within the electronic device or may be collected by a collection module externally connected to the electronic device.

[0082] Furthermore, in this embodiment of the present application, the following two aspects may be considered when the competing obstacle screening module determines the competing object based on the recognition information:

[0083] In a first aspect, the competing obstacle screening module may determine at least one object based on the acquired recognition information, and predict the moving path of the at least one object.Then, the competing obstacle screening module determines the object whose moving path of the at least one object is in conflict with the moving path of the own vehicle as a competing object.The path conflict in this embodiment of the present application is used to indicate the case where the object is currently in a moving state, and the future moving path of the object collides with the moving path of the own vehicle.

[0084] For example, as shown in FIG. 6, assume that the moving path of the self-vehicle is straight, the competing obstacle screening module determines that two objects, object 1 and object 2, are in the current range based on the acquired recognition information, and object 1 and object 2 are in a moving state. The competing obstacle screening module may generate a predicted path corresponding to object 1 and a predicted path corresponding to object 2 based on the road topology and the movement status. The predicted path of object 1 conflicts with the moving path of the self-vehicle, object 2 is in a state of moving in the same direction as the self-vehicle, and the predicted path of object 2 does not conflict with the moving path of the self-vehicle. In this case, the competing obstacle screening module determines object 1 as a competing object, that is, the self-vehicle is involved in a conflict with object 1 in the moving process. The subsequent movement judgment of the self-vehicle needs to be made based on the predicted path of object 1.

[0085] In this embodiment of the present application, whether the predicted path of the object conflicts with the moving path of the ego-vehicle may be determined in two ways:

[0086] Method 1: When the predicted path of object A intersects with the path of the host vehicle and object A is not behind the host vehicle (i.e., the object is a non-following object), the competing obstacle screening module can determine object A as a competing object, taking into account that the predicted path of object A conflicts with the path of the host vehicle at a future time point starting from the current time point.

[0087] As an example, assume that the current time is 10:08:20. The host vehicle predicts the future path of object A, and recognizes that the path of the host vehicle and the path of object A will intersect at 10:08:25 in the future when the host vehicle and object A travel based on the predicted future path, by referring to the future path of the host vehicle. Therefore, the conflicting obstacle screening module determines that the predicted path of object A will conflict with the path of the host vehicle in the future.

[0088] Method 2: If the narrow angle between the predicted path of object A and the movement path of the host vehicle is less than the threshold narrow angle, the competing obstacle screening module can determine object A as a competing object, taking into account that the predicted path of object A conflicts with the movement path of the host vehicle at a future time starting from the current time.

[0089] As an example, assume that the current time is 11:09:32. The host vehicle predicts the future moving path of the object A, and by referring to the future moving path of the host vehicle, recognizes that when the host vehicle and the object A travel based on the predicted future moving path, the narrow angle between the moving path of the host vehicle and the moving path of the object A is less than the threshold narrow angle at 11:09:36 in the future (for example, the threshold narrow angle may be 60 degrees). Therefore, the conflicting obstacle screening module determines that the predicted path of the object A will conflict with the moving path of the host vehicle in the future.

[0090] In a second aspect, the competing obstacle screening module may determine at least one object based on the acquired recognition information, and may predict the motion intention of one object.Then, the competing obstacle screening module determines the object among the at least one object whose motion intention has a conflict with the moving path of the vehicle as a competing object.The intention conflict in this embodiment of the present application is used to indicate the case where the object is currently in a stationary state, will be in a moving state in the future, and the future motion intention of the object conflicts with the moving path of the vehicle.

[0091] For example, assume that the moving path of the self-vehicle is a straight line, as shown in FIG. 7. Based on the acquired recognition information, the competing obstacle screening module determines that two objects, object 1 and object 2, are in the current recognition range of the self-vehicle, and currently object 1 and object 2 are stationary. Based on the acquired recognition information, the competing obstacle screening module further predicts that the movement intention of object 1 is to move from point A to point B, which may not conflict with the moving path of the self-vehicle, and the movement intention of object 2 is to turn right, which may conflict with the moving path of the self-vehicle at a future time. Therefore, the competing obstacle screening module determines object 2 as a competing object, i.e., the self-vehicle is involved in a conflict with object 2 in the moving process, and the subsequent movement judgment of the self-vehicle needs to be made based on the predicted path of object 2.

[0092] In this embodiment of the present application, whether the object's motion intention conflicts with the movement path of the ego-vehicle may be determined in the following two ways.

[0093] Method 1: When object A is currently in a stationary state, the competing obstacle screening module determines the movement intention of object A based on the recognition information, and predicts that the future movement path of object A will intersect with the movement path of the host vehicle based on the movement intention of object A. In this case, the competing obstacle screening module can consider that the movement intention of object A will compete with the movement path of the host vehicle at a future time point starting from the current time point, and the competing obstacle screening module determines object A as a competing object.

[0094] Method 2: When object A is currently in a stationary state, the competing obstacle screening module determines the movement intention of object A based on the recognition information, and predicts that the narrow angle between the future movement path of object A and the movement path of the host vehicle is less than the threshold narrow angle based on the movement intention of object A. In this case, the competing obstacle screening module can consider that the movement intention of object A conflicts with the movement path of the host vehicle at a future time point starting from the current time point, and the competing obstacle screening module determines object A as a competing object.

[0095] The conflicting object status tracking module is configured to continuously observe the motion status of the conflicting object, including the left-right status feature and the front-rear status feature, in the process of the conflict between the ego-vehicle and the conflicting object, and record key information, thereby providing input for subsequently determining the match or mismatch between the actual movement path of the conflicting object and the predicted movement path of the judgment solution.

[0096] The conflict adjudication module is configured to generate adjudication solutions.

[0097] Alternatively, the conflict decision module may include a policy feasible field generation module and a decision state machine module.

[0098] The policy feasibility field generation module is configured to generate a judgment solution set based on the obtained recognition information.

[0099] In particular, as shown in Fig. 8, the policy feasible field generation module may refer to longitudinal feature information such as road speed limit, acceleration change Jerk value, and conflict object type to generate a longitudinal acceleration sampling range based on the acquired recognition information, and may refer to feature information such as road boundary, stationary obstacles, and vehicle kinematics to generate a left-right offset sampling range. The policy feasible field generation module combines the left-right position offset sampling range and the longitudinal acceleration sampling range to form a sampling space, and the formed sampling space is a possible space of all policies. The policy feasible field generation module may obtain a set of all judgment solutions based on the formed sampling space.

[0100] The policy feasible field generation module may divide the judgment solution set into at least one judgment solution subset based on a category of an action pair between the ego vehicle and the conflict target included in each judgment solution in the judgment solution set. Further, the policy feasible field generation module may mark every action pair between the ego vehicle and the conflict target in the judgment solution set with a judgment indicator, and group clusters with matching indicators into judgment solution subsets, each judgment solution subset including at least one judgment solution.

[0101] In this way, the decision solution set can be divided into multiple decision solution subsets. For example, in some scenes, the policy feasible field generation module may divide the decision solution set into four decision solution subsets: avoid and rush through, yield and keep, avoid and rush through and keep, and avoid and yield.

[0102] A target judgment solution in the embodiment of the present application is a judgment solution with the maximum judgment benefit in a corresponding judgment solution subset. For example, assume that a judgment solution subset A includes three judgment solutions, namely, judgment solution 1, judgment solution 2, and judgment solution 3, and the judgment benefit of judgment solution 3 exceeds the judgment benefit of judgment solution 2 and the judgment benefit of judgment solution 1. In this case, judgment solution 1 is a target judgment solution in the judgment solution subset A.

[0103] In addition to the above, in this embodiment of the present application, the policy feasibility field generation module may determine a judgment benefit corresponding to the judgment solution based on a first calculation method (e.g., a benefit evaluation method provided in this embodiment of the present application used to calculate the benefit corresponding to the judgment solution).

[0104] The benefit evaluation method provided in this embodiment of the present application, used to calculate the benefit corresponding to a judgment solution, may include several judgment measures such as the safety, comfort, passability, right-of-way, offset, and judgment result history of the judgment solution.

[0105] For example, the judgment benefit of the benefit evaluation method may be determined based on the judgment cost. A small cost indicates that the judgment benefit is small and the judgment solution is more likely to be used as the optimal judgment solution. The specific evaluation magnitude and description are as follows:

[0106] (1) Safety cost: The safety cost in this embodiment of the present application is mainly a judgment cost obtained based on the distance between the host vehicle and a competing vehicle in the process of deriving the movement path of the host vehicle and the movement path of a competing vehicle.

[0107] A small distance between the host vehicle and a competing vehicle indicates low safety. In other words, a small distance between the host vehicle and a competing vehicle indicates that the safety cost of the host vehicle is high and the benefit of making a decision based on the host vehicle's safety is small.

[0108] Furthermore, in this embodiment of the present application, a minimum distance between the host vehicle and the competing vehicle may be further set. In the process in which the distance between the host vehicle and the competing vehicle gradually approaches the minimum distance, the safety cost gradually increases and the benefit of the safety-based decision gradually decreases. When the distance between the host vehicle and the competing vehicle does not exceed the minimum distance, the safety cost reaches a maximum and the benefit of the safety-based decision reaches a minimum.

[0109] In addition to the above, in this embodiment of the present application, a maximum distance between the self-vehicle and the competing vehicle used to calculate the safety cost may be further set. When the distance between the self-vehicle and the competing vehicle is not less than the maximum distance, the safety cost reaches a minimum and the benefit of the safety-based decision reaches a maximum.

[0110] For example, as shown in FIG. 9(a), the horizontal axis represents the distance between the host vehicle and the competing vehicle, point A on the horizontal axis represents the minimum distance between the host vehicle and the competing vehicle set in this embodiment of the present application, point B on the horizontal axis can be said to represent the maximum distance between the host vehicle and the competing vehicle set in this embodiment of the present application which is used to calculate the safety cost, and the vertical axis represents the safety cost.

[0111] When the distance between the own vehicle and the competing vehicle is not less than B, the safety cost is minimized and the benefit of a safety-based decision is maximized. When the distance between the own vehicle and the competing vehicle is between point A and point B, as the distance between the own vehicle and the competing vehicle approaches point A, the safety cost increases and the benefit of a safety-based decision decreases, and as the distance between the own vehicle and the competing vehicle approaches point B, the safety cost decreases and the benefit of a safety-based decision increases. When the distance between the own vehicle and the competing vehicle does not exceed A, the safety cost is maximized and the benefit of a safety-based decision is minimized.

[0112] In this embodiment of the present application, point A and point B shown in (a) of Fig. 9 may be obtained based on actual sensation. For example, in this embodiment of the present application, point A shown in (a) of Fig. 9 may be 0.2m to 0.5m (for example, point A is 0.5m), and point B may be 1.0m to 3.0m (for example, point B is 2.5m).

[0113] (2) Comfort cost: The comfort cost in this embodiment of the present application is a judgment cost obtained mainly based on the difference between the current acceleration and the expected acceleration of the host vehicle.

[0114] A small difference between the vehicle's current acceleration and the predicted acceleration indicates that the vehicle's comfort cost is small and that the benefit of making comfort-based decisions is great.

[0115] Furthermore, in this embodiment of the present application, a maximum difference between the current acceleration and the expected acceleration of the vehicle may be further set. When the difference between the current acceleration and the expected acceleration of the vehicle is closer to the maximum difference, the comfort cost is larger and the benefit of the comfort-based decision is smaller.

[0116] For example, as shown in FIG. 9(b), the horizontal axis represents the difference between the current acceleration and the expected acceleration of the host vehicle. For example, the current acceleration of the host vehicle is 1m / S^2, and the expected acceleration is 4m / S^2. In this case, the difference between the current acceleration and the expected acceleration of the host vehicle is 3m / S^2. Point A on the horizontal axis represents the maximum difference between the current acceleration and the expected acceleration of the host vehicle set in this embodiment of the present application.

[0117] If the difference between the current acceleration and the expected acceleration of the host vehicle is closer to point A, the comfort cost is greater and the benefit of a comfort-based decision is smaller. If the difference between the current acceleration and the expected acceleration of the host vehicle is farther from point A, the comfort cost is smaller and the benefit of a comfort-based decision is greater. If the difference between the current acceleration and the expected acceleration of the host vehicle is not less than the acceleration value corresponding to point A, the comfort cost is maximized and the benefit of a comfort-based decision is minimized.

[0118] In this embodiment of the present application, point A shown in FIG. 9(b) may be obtained based on actual sensation.

[0119] For example, in this embodiment of the present application, point A shown in FIG. 9(b) may be 5m / S^2 to 7m / S^2 (for example, point A is 6m / S^2).

[0120] (3) Passability cost: In this embodiment of the present application, the passability cost is a judgment cost that is obtained mainly based on the speed difference between the current speed of the host vehicle and the speed at which the host vehicle reaches a conflict point (e.g., a point at which the host vehicle's path may intersect with the path of a conflicting object).

[0121] A large speed difference (current speed minus the speed at the conflict point) indicates a high passability cost and a small benefit of making a decision based on the passability of the own vehicle.

[0122] Moreover, in this embodiment of the present application, a maximum value of the speed difference may be further set. When the speed difference is closer to the maximum value, the traversability cost is larger and the benefit of traversability-based decision is smaller.

[0123] For example, as shown in FIG. 9(c), the horizontal axis represents the speed difference, and point A on the horizontal axis represents the maximum value of the speed difference set in this embodiment of the present application. If the speed difference is closer to point A, the traversability cost is larger and the benefit of the traversability-based decision is smaller. If the speed difference is farther from point A, the traversability cost is smaller and the benefit of the traversability-based decision is larger. If the speed difference is not less than the speed difference corresponding to point A, the traversability cost is maximized and the benefit of the traversability-based decision is minimized.

[0124] In this embodiment of the present application, point A shown in FIG. 9(c) may be obtained based on actual sensation.

[0125] For example, in this embodiment of the present application, point A shown in FIG. 9(c) may be 6 m / s to 10 m / s (eg, point A is 7 m / s).

[0126] (4) Right-of-way cost: In this embodiment of the present application, the right-of-way cost is a judgment cost obtained based on the right-of-way status of the own vehicle or a competing object with reference to the current acceleration of the own vehicle.

[0127] In the present embodiment of the present application, the calculation is mainly performed based on the side having the strong right of way, and the right of way cost penalty is performed on the side having the right of way. For example, when the host vehicle has the strong right of way, the right of way cost penalty is performed when the host vehicle slows down.

[0128] It can be seen that the priority right in this embodiment of the present application is a relative relationship between the vehicle itself and the competing object. If the priority right of the vehicle itself is stronger than that of the competing object, the vehicle with the stronger priority right will naturally not change its movement status as much as possible in the interaction competition process. Therefore, when the acceleration of the vehicle itself decreases and the vehicle itself decelerates, a relatively strong priority right cost penalty is naturally imposed on the change in the movement status of the vehicle itself. In other words, when the priority right of the vehicle itself is stronger than that of the competing object, the current acceleration of the vehicle itself is small, which indicates that the priority right cost of the vehicle itself is large and the benefit of the judgment based on the priority right of the vehicle itself is small.

[0129] Furthermore, in this embodiment of the present application, a minimum value of the current acceleration of the host vehicle may be further set. When the current acceleration of the host vehicle is closer to the minimum value, the right-of-way cost is larger and the benefit of the right-of-way-based decision is smaller.

[0130] For example, as shown in Fig. 9(d), ACC represents acceleration, and the strong priority of the ego vehicle is the restrictive premise. For example, the strong priority of the ego vehicle is used as the prerequisite input, and the priority cost is further determined based on the current priority relationship between the ego vehicle and the competing target.

[0131] The horizontal axis represents the current acceleration of the host vehicle, and point A on the horizontal axis represents the minimum difference value of the current acceleration of the host vehicle set in this embodiment of the present application. If the current acceleration of the host vehicle is closer to point A, the right of way cost is larger and the benefit of the decision based on the right of way is smaller. If the current acceleration of the host vehicle is farther from point A, the right of way cost is smaller and the benefit of the decision based on the right of way is larger. If the current acceleration of the host vehicle does not exceed the acceleration value corresponding to point A, the right of way cost is maximized and the benefit of the decision based on the right of way is minimized.

[0132] In this embodiment of the present application, point A shown in FIG. 9(d) may be obtained based on actual sensation.

[0133] As an example, in this embodiment of the present application, point A shown in (d) of Figure 9 may be (-1)m / S^2 to (-3)m / S^2 (for example, point A is (-1)m / S^2).

[0134] (5) Offset cost: The offset cost in this embodiment of the present application is mainly a judgment cost obtained based on the position offset of the host vehicle relative to the reference line. It can be seen that a long offset distance of the host vehicle relative to the reference line indicates a large offset cost value of the host vehicle, and a small benefit of judgment based on the host vehicle offset.

[0135] Alternatively, the reference line in this embodiment of the present application may be the centre line of the road on which the vehicle is located.

[0136] 10, in this embodiment of the present application, there may be two offset regions, for example, region A and region B. In this embodiment of the present application, region A may be referred to as a soft boundary and region B may be referred to as a hard boundary.

[0137] From the contents of FIG. 10, it can be seen that the soft boundary set in this embodiment of the present application may be the area of ​​the driving lane where the vehicle is located. In other words, when the host vehicle is offset with respect to the reference line, but the host vehicle maintains driving in the driving lane, it can be seen that the host vehicle is always within the soft boundary. The hard boundary set in this embodiment of the present application may be the area outside the driving lane. In other words, it can be seen that when the host vehicle is offset with respect to the reference line and the host vehicle deviates from the driving lane, the host vehicle enters the hard boundary.

[0138] In addition to the above, when the vehicle moves away from the driving lane, the offset of the vehicle relative to the reference line is larger, there are relatively many risk factors in the driving process, and the safety is relatively low. Therefore, in this embodiment of the present application, when the vehicle moves away from the driving lane, that is, when the vehicle is in the hard boundary, the penalty cost may be increased. For example, in this embodiment of the present application, different gradient values ​​for calculating offset cost may be used in the calculation based on different regions. The gradient value set at the soft boundary and used to calculate offset cost may be relatively small, and the gradient value set at the hard boundary and used to calculate offset cost may be relatively large.

[0139] 9(e), the horizontal axis represents the current offset distance between the vehicle and the reference line, point -A on the horizontal axis represents the leftmost position of the soft boundary set in this embodiment of the present application, point 0 represents the reference line position, point A represents the rightmost position of the soft boundary set in this embodiment of the present application, point -B represents the leftmost position of the hard boundary set in this embodiment of the present application, and point B represents the rightmost position of the hard boundary set in this embodiment of the present application. The offset cost of the vehicle relative to the reference line at the soft boundary is calculated using gradient 1, and the offset cost of the vehicle relative to the reference line at the hard boundary is calculated using gradient 2, where gradient 2 exceeds gradient 1.

[0140] When the offset distance between the host vehicle and point 0 of the reference line is larger, the offset cost is larger and the benefit of the offset-based decision is smaller. When the offset distance between the host vehicle and point 0 of the reference line is not less than the distance corresponding to point C, the offset cost reaches a maximum and the benefit of the offset-based decision reaches a minimum.

[0141] In this embodiment of the present application, point A, point B, gradient 1 and gradient 2 shown in FIG. 9(e) may be obtained based on actual sensation.

[0142] For example, in this embodiment of the application, point A shown in (e) of FIG. 9 may be between 1.5m and 2.5m (e.g., if point A is 2m, the corresponding point -A is -2m), and point B may be between 1.9m and 2.9m (e.g., if point B is 2.9m, the corresponding point -B is -2.9m).

[0143] (6) Judgment result history cost: The judgment result history cost in this embodiment of the present application may also be referred to as inter-frame related cost, which is mainly the judgment cost of the current frame obtained based on the judgment solution of the last frame of the own vehicle.

[0144] It can be seen that for the same competing object, if the last frame of the own vehicle is a rush-passing one and the next frame of the own vehicle is a yielding one, the speed change of the own vehicle is relatively large, so that the driving fluctuation is relatively large and the comfort is reduced. Therefore, in order to effectively ensure that the driving fluctuation is relatively small, the fluctuation is mitigated, and the driving is more stable, when the own vehicle rushes through and takes the lead, the own vehicle maintains the rush-passing state as much as possible.

[0145] For example, as shown in (f) of FIG. 9, when the Kth frame is a rushing frame and the (K+1)th frame is also a rushing frame, the judgment result history cost is relatively small and the judgment profit based on the judgment result history is relatively large. In contrast, when the Kth frame is a rushing frame and the (K+1)th frame is a yielding frame, the judgment result history cost is relatively large and the judgment profit based on the judgment result history is relatively small.

[0146] Furthermore, the electronic device of this embodiment of the present application may combine the costs of the aforementioned evaluation magnitudes based on certain weighting ratios to obtain multiple magnitudes of judgment costs.

[0147] In some embodiments, the benefit evaluation scheme may be determined primarily based on the judgment result history, i.e., may depend primarily on the judgment result history (e.g., the weight of the judgment result history cost is set to 80%-90%), and the vehicle's driving style may be determined based on the judgment result history, where the driving style is, for example, careful, moderate or rough.

[0148] In some embodiments, the weights corresponding to the magnitude of the evaluation of the benefit evaluation method of this embodiment of the present application are: Safety: 10000, Passability: 6000, Comfort: 500, Right of Way: 1000, Offset: 2000 and Danger Area: 1000 is set to.

[0149] Therefore, the judgment cost obtained in this embodiment of the present application is Safety cost x 10,000 + Passability cost x 6,000 + Comfort cost x 500 + Right of way cost x 1,000 + Offset cost x 2,000 + Danger area cost x 1,000 may be also possible.

[0150] Furthermore, the weights corresponding to the magnitude of the evaluation of the benefit evaluation method may be different for different driving environments.

[0151] As an example, the correspondence between the driving environment and the weight of the evaluation magnitude may be as shown in Table 1.

[0152] [Table 1]

[0153] If the current driving environment of the vehicle is an urban road, the policy feasibility field generation module may determine, based on the contents of Table 1, that the weights corresponding to the evaluation magnitude of the benefit evaluation method used are safety cost x 10,000, passability cost x 6,000, comfort cost x 500, right of way cost x 1,000, offset cost x 2,000 and dangerous area cost x 1,000.

[0154] The decision state machine module may classify the decision solution subsets in the decision solution set into a hierarchy based on the decision solution benefit of each decision solution subset.

[0155] For example, the judgment solution subsets of the judgment solution set are classified into two hierarchies, the first hierarchies and the second hierarchies. The judgment profit of the target judgment solution corresponding to each judgment solution subset in the first hierarchies is not less than the threshold profit, and the judgment profit of the target judgment solution corresponding to each judgment solution subset in the second hierarchies is less than the threshold profit. From the viewpoint of the profit of the judgment solutions, all the profits of the judgment solutions in the first hierarchies exceed the profits of the judgment solutions in the second hierarchies. That is, at the present time, the judgment solutions in the first hierarchies are "better" than the judgment solutions in the second hierarchies.

[0156] Furthermore, if the decision state machine module determines that the difference in decision benefits of the optimal decision solutions of any two decision solution subsets in the first hierarchy does not exceed the threshold profit difference, then it enters the first interaction stage. In this case, the interaction relationship between the ego vehicle and the competing target is unknown. The optimal decision solution of each decision solution subset is the decision solution with the maximum profit in the decision solution subset.

[0157] In some cases, if the judgment state machine module determines, after performing hierarchical division and before entering the first interaction stage, that the judgment benefit of the judgment solution in the first hierarchy clearly exceeds the judgment benefit of all other judgment solutions in the first hierarchy, the judgment state machine module may directly enter the second interaction stage, in which case the interaction relationship between the ego-vehicle and the competing target is clear.

[0158] As an example, assume that the threshold profit is 0.76 and the threshold profit difference is 0.05. In the example scene (e.g., the ego vehicle is going straight and the vehicles are merging at the intersection), the decision solution set includes four decision solution subsets: decision solution subset 1 (corresponding to the ego vehicle rushing through while avoiding), decision solution subset 2 (corresponding to the ego vehicle rushing through without avoiding), decision solution subset 3 (corresponding to the ego vehicle yielding without avoiding), and decision solution 4 (corresponding to the ego vehicle yielding while avoiding).

[0159] The optimal judgment solution in the judgment solution subset 1 is the judgment solution 1, with a corresponding judgment profit of 0.8, the optimal judgment solution in the judgment solution subset 2 is the judgment solution 2, with a corresponding judgment profit of 0.73, the optimal judgment solution in the judgment solution subset 3 is the target judgment solution 3, with a corresponding judgment profit of 0.77, and the optimal judgment solution in the judgment solution subset 4 is the target judgment solution 4, with a corresponding judgment profit of 0.69. For each judgment solution subset, the optimal judgment solution of the judgment solution subset can be considered to represent the judgment solution subset. In other words, each judgment solution subset (e.g., the ego vehicle rushes through while avoiding) may include multiple judgment solutions, but one of the multiple judgment solutions may have the maximum profit, and this judgment solution may represent the judgment solution subset as the optimal judgment solution of the judgment solution subset.

[0160] First, since the judgment profits of judgment solution 1 and judgment solution 3 exceed the threshold profit of 0.76, and the judgment profits of judgment solution 2 and judgment solution 4 are less than the threshold profit, the judgment state machine module classifies judgment solution subset 1 and judgment solution subset 3 into the first tier, and classifies judgment solution subset 2 and judgment solution subset 4 into the second tier.

[0161] Then, the judgment state machine module only needs to determine through comparison whether the difference in judgment profit of the optimal judgment solution of the first hierarchical level is less than the threshold profit difference of 0.05. The difference between the judgment profit of the optimal judgment solution 1 and the judgment profit of the optimal judgment solution 3 is 0.03, which is less than the threshold profit difference of 0.05. Therefore, at present, the judgment state machine module cannot determine which judgment solution (i.e., the ego-vehicle rushes past while avoiding or the ego-vehicle avoids while yielding) is more suitable for this conflict, i.e., it cannot determine whether to use the optimal judgment solution 1 or the optimal judgment solution 3 as the driving policy. In this case, it can be seen that the interaction relationship between the ego-vehicle and the conflicting object is unknown and needs to be entered into a first interaction stage, and in the first interaction stage, the ego-vehicle and the conflicting object may jump between multiple corresponding various judgment solutions.

[0162] Furthermore, if it is determined that the interaction between the host vehicle and the competing object satisfies the condition for entering the second interaction stage, the judgment state machine module enters the second interaction stage from the first interaction stage. In this case, the interaction relationship between the host vehicle and the competing object is clear.

[0163] The decision state machine module may determine whether the conditions for entering the second interaction phase are met, i.e., whether to enter the second interaction phase from the first interaction phase, based on the following two aspects:

[0164] In a first aspect, the decision state machine module may determine whether a condition for entering the second interaction stage is met based on a relationship between a decision profit difference between the decision solution with the first largest decision profit and the decision solution with the second largest decision profit in the decision solution set and a threshold profit difference.

[0165] In one case, if the decision profit difference between the decision solution with the first largest decision profit and the decision solution with the second largest decision profit in the decision solution set is not less than the threshold profit difference, the decision state machine module determines that the condition for entering the second interaction phase is met.

[0166] If the decision profit difference between the decision solution with the highest decision profit (e.g., decision solution A) and the decision solution with the second highest decision profit (e.g., decision solution B) in the decision solution set is less than the threshold profit difference, it can be seen that this indicates that the profit difference between the competitive interaction using decision solution A and the competitive interaction using decision solution B is relatively small. In this case, the decision state machine module cannot determine whether decision solution A or decision solution B is more suitable for the current conflict, and the decision state machine module may determine that the current interaction relationship between the ego-vehicle and the conflict target is unknown and needs to be put into the first interaction stage.

[0167] In another case, if the decision profit difference between the decision solution with the first largest decision profit and the decision solution with the second largest decision profit in the decision solution set is less than the threshold profit difference, the decision state machine module determines that the condition for entering the second interaction stage is not met.

[0168] If the judgment profit difference between the judgment solution with the highest judgment profit (e.g., judgment solution A) and the judgment solution with the second highest judgment profit (e.g., judgment solution B) in the judgment solution set is not less than the threshold profit difference, for example, if the judgment profit difference between the judgment profit of judgment solution A and the judgment profit of judgment solution B exceeds the threshold profit difference, it can be seen that this indicates that judgment solution A is suitable to carry out this competitive interaction. In this case, the judgment state machine module can determine that judgment solution A is more suitable for this competition, so the judgment state machine module may determine that the current interaction relationship between the ego-vehicle and the competing target is clear and needs to be entered into the second interaction stage.

[0169] As an example, assume that the threshold profit difference is 0.03 and the first tier includes optimal decision solution 1 and optimal decision solution 2. Assume that the decision state machine module currently determines that the decision profit of optimal decision solution 1 is 0.81 and the decision profit of optimal decision solution 2 is 0.77. The decision profit of optimal decision solution 1 exceeds the decision profit of optimal decision solution 2 by 0.04, and the decision profit difference exceeds the threshold profit difference of 0.03.

[0170] In this case, the decision state machine module may determine that the decision benefit of the currently optimal decision solution 1 in the first tier clearly exceeds the decision benefit of the optimal decision solution 2. The decision state machine module may determine that the optimal decision solution 1 is more suitable for the current conflict and is entered into a second interaction stage, where the optimal decision solution 1 is selected as the second decision solution for running.

[0171] In a second aspect, a decision state machine module may determine whether a condition for entering a second interaction phase is met based on a comparison result between an actual movement path of a conflicting object at a given time point and a predicted path of the conflicting object using the first decision solution.

[0172] In one case, the decision state machine module determines that a condition for entering a second interaction phase is met if the decision state machine module determines that the actual movement path of the conflicting object at a given time has a relatively high consistency with the movement path of the conflicting object at a given time predicted using the first decision solution, where the first decision solution used in the first interaction phase is the same as the second decision solution used in the second interaction phase.

[0173] If the judgment state machine module determines that the actual movement path of the conflicting object at a given time is consistent with the movement path of the conflicting object at a given time predicted using a first judgment solution (e.g., judgment solution A), this indicates that the movement path of the conflicting object is completely compatible with the prediction of the judgment solution, which means that the judgment solution A currently used in the conflict interaction can be used. In this case, the judgment state machine module can determine that the judgment solution A is suitable for the conflict. Therefore, the judgment state machine module may determine that the second interaction phase is entered because the current interaction relationship between the ego-vehicle and the conflicting object is clear. Since the judgment state machine module determines that the judgment solution A is suitable for the current conflict when entering the second interaction phase, the judgment state machine module may continue to use the judgment solution A as the second judgment solution for executing the second interaction phase.

[0174] In another case, if the decision state machine module determines that the actual movement path of the conflicting object at a given time does not match the movement path of the conflicting object at a given time predicted using the first decision solution, the decision state machine module determines that the condition for entering the second interaction phase is met, where the first decision solution used in the first interaction phase is different from the second decision solution used in the second interaction phase.

[0175] If the decision state machine module determines that the actual movement path of the conflicting object at a given time does not match the movement path of the conflicting object at a given time predicted using the first decision solution (e.g., decision solution A), it can be seen that this indicates that the movement path of the conflicting object is completely unexpected, which means that the decision solution A currently used for the conflict interaction is inapplicable. In this case, the decision state machine module needs to adjust the decision solution A used for the conflict interaction to a decision solution (e.g., decision solution B) that is more suitable for the current conflict. The decision solution B may be obtained from the decision solution set. For example, if the ego-vehicle currently detects that the policy of rushing through without avoiding is no longer applicable (the conflicting object may not yield as expected, so rushing through without slowing down), a more suitable decision solution (e.g., yielding while avoiding) may be obtained from the decision solution set. After the decision solution B is determined, the decision state machine module may determine that the current interaction relationship between the ego-vehicle and the conflicting object is clear and enters the second interaction phase.

[0176] Additionally, the decision state machine module may further combine the first aspect with the second aspect, which together are used to determine whether a condition for entering the second interaction phase is met.

[0177] In the case where the first and second modes are combined to determine whether the condition for entering the second interaction phase is satisfied, and the result of determining whether the condition for entering the second interaction phase is satisfied based on the first mode is inconsistent with the result of determining whether the condition for entering the second interaction phase is satisfied based on the second mode, the result of determining whether the condition for entering the second interaction phase is satisfied based on the second mode is used, i.e., the priority of the second mode exceeds the priority of the first mode. The reason for this setting is that compatibility / inconsistency jumping is usually caused by competing objects that do not follow traffic rules or general driving behavior, and compatibility / inconsistency jumping may be contradictory to cost evaluation, and it should be noted that the priority of compatibility / inconsistency jumping exceeds the priority of benefit-based jumping.

[0178] For example, the decision state machine module determines based on the second aspect that the actual movement path of the conflicting object at a given time point is not consistent with the movement path of the conflicting object at a given time point predicted using the first decision solution, and needs to be put into the second interaction phase. In this case, the decision state machine module recognizes through the determination based on the first aspect that the decision benefit difference between the decision solution A having the first largest decision benefit and the decision solution B having the second largest decision benefit in the decision solution set is less than the threshold benefit difference, and does not need to be put into the second interaction phase. Since the priority of the second aspect exceeds the priority of the first aspect, the decision state machine module mainly determines that the second interaction phase needs to be put into the second interaction phase based on the result determined in the second aspect.

[0179] In addition to the above, in this embodiment of the present application, in addition to the two aspects mentioned above, whether the conditions for entering the second interaction stage are met may be further determined in another manner or in combination with another manner.

[0180] For example, the decision state machine module may determine whether to enter a second interaction phase based on an interaction duration of a first interaction phase.

[0181] Separately from the above, after determining to enter the first interaction phase, the decision state machine module starts a corresponding timer, the timer's time duration being the threshold duration, and when the decision state machine module determines that the timer has expired, the decision state machine module determines that the condition for entering the second interaction phase is met and the second interaction phase is entered.

[0182] As an example, assume the threshold period is 10 s. If the decision state machine module determines that the period of time spent in the first interaction phase is greater than 10 s, the decision state machine module enters the second interaction phase.

[0183] In another example, the decision state machine module may determine whether to enter the second interaction phase based on the distance between the ego vehicle and the conflicting object.

[0184] Alternatively, after the decision state machine module determines to enter the first interaction phase, the distance between the host vehicle and the conflicting object may decrease while continuing to interact with each other. If the decision state machine module determines that the distance between the host vehicle and the conflicting object is less than a threshold distance, the decision state machine module determines that a condition for entering a second interaction phase is met and to enter the second interaction phase.

[0185] Alternatively, the decision state machine module may determine a match or mismatch based on the left-right and front-back motion sequences of the competing objects.

[0186] According to the second aspect, when the judgment state machine module determines whether the predicted path of the conflicting object at a given time point is consistent with the actual path of the conflicting object at a given time point, the judgment state machine module may determine the future status (i.e., predicted path) of the conflicting object at a given time point based on the continuous status change history and based on the predicted action sequence of the conflicting object according to the judgment solution selected for each interaction time point and the actual action sequence of the conflicting object, and the left-right and front-back directions may be distinguished in the decision process.

[0187] As an example, the competing object is a bicycle.,The predicted acceleration sequence and the actual acceleration sequence are,processed using statistical data processing methods to obtain the,forward / backward motion sequence shown in Fig. 11(a).

[0188] According to the contents shown in FIG. 11(a), it can be seen that a1 to a3 are predicted acceleration changes of the competing object in the predicted period determined by the judgment state machine module, b1 to b3 are actual acceleration changes of the competing object in the current period, and b4 to b6 are actual acceleration changes of the competing object in the predicted period. If the difference between a1 to a3 and b4 to b6 exceeds the threshold range, the judgment state machine module determines that the tracking result of the competing object at the predetermined time is not consistent with the movement path of the competing object at the predetermined time predicted using the first judgment solution. If the difference between a1 to a3 and b4 to b6 fluctuates within the threshold range, the judgment state machine module determines that the tracking result of the competing object at the predetermined time is consistent with the movement path of the competing object at the predetermined time predicted using the first judgment solution.

[0189] As an example, the competing object is a bicycle. A statistical data processing method is used to process the predicted distance sequence and the actual distance sequence to obtain the lateral motion sequence shown in Fig. 11(b). Since the lateral motion change is usually irreversible, the lateral match / mismatch decision is influenced by the decision result history.

[0190] According to the contents shown in FIG. 11(b), it can be seen that c1 to c3 are the predicted distance changes of the conflicting objects in the predicted period determined by the judgment state machine module, d1 to d3 are the actual distance changes of the conflicting objects in the current period, and d4 to d6 are the actual distance changes of the conflicting objects in the predicted period. If the difference between c1 to c3 and d4 to d6 exceeds the threshold range, the judgment state machine module determines that the tracking result of the conflicting objects at a predetermined time does not match the movement path of the conflicting objects at a predetermined time predicted using the first judgment solution. If the difference between c1 to c3 and d4 to d6 fluctuates within the threshold range, the judgment state machine module determines that the tracking result of the conflicting objects at a predetermined time matches the movement path of the conflicting objects at a predetermined time predicted using the first judgment solution.

[0191] The motion planning module needs to perform corresponding actions based on the decision solutions of the first interaction stage and the second interaction stage.

[0192] The system architecture and application scenarios described in the embodiments of the present application are intended to more clearly explain the technical solutions of the embodiments of the present application, and do not constitute limitations on the technical solutions provided in the embodiments of the present application.

[0193] In addition, those skilled in the art can imagine that the technical solutions provided in the embodiments of the present application can also be applied to similar technical problems when vehicle architecture evolves and new service scenarios emerge. It should be understood that Figures 2 to 5 are merely simplified schematic diagrams of examples for easy understanding. The system architecture may further include other devices or further include other unit modules.

[0194] Hereinafter, the method provided in the embodiment of the present application will be described with reference to Fig. 12. The method may be performed by an electronic device. The electronic device may include a processing device and a communication device. The processing device may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a central processor unit (CPU), etc. The electronic device may include any one or more structures shown in Figs. 2 to 5.

[0195] When the intelligent driving method is implemented, the processing module 210 shown in Fig. 2, the processor shown in Fig. 4, or the competing obstacle screening module, the competing object status tracking module, the competing judgment module and the motion planning module shown in Fig. 5 may be used to implement the processing operations of the method provided in the embodiment of the present application. The collection module 230 shown in Fig. 2, the collection device shown in Fig. 4, etc. may collect the recognition information. The interaction includes, but is not limited to, obtaining the recognition information of the vehicle itself and the recognition information of the competing object.

[0196] S1200: The electronic device obtains a set of judgment solutions for the host vehicle based on the recognition information.

[0197] S1201: The electronic device obtains, from a set of judgment solutions, a first judgment solution to be used in a first interaction stage between the host vehicle and a competing target.

[0198] S1202: In the process of controlling the own vehicle to which the electronic device belongs to run based on the first judgment solution, the electronic device detects that the own vehicle and the competing object satisfy a condition to enter a second interaction stage.

[0199] S1203: The electronic device obtains, from the set of judgment solutions, a second judgment solution to be used in a second interaction stage between the host vehicle and the conflict target.

[0200] S1204: The electronic device controls the host vehicle to travel based on the second determination solution.

[0201] It is understood that the division of the intelligent driving method of the present application into a first interaction stage and a second interaction stage does not constitute a limitation to the present application.

[0202] For example, in the present application, the judgment process may be divided into three interaction stages based on the actual situation.

[0203] Alternatively, in this embodiment of the present application, the first interaction stage may be referred to as a crisis state.

[0204] The crisis state in this embodiment of the present application can be understood as one in which the interaction relationship between the host vehicle and the conflicting object is unknown, i.e., the determination solution of the interaction between the host vehicle and the conflicting object may change at any time.

[0205] Alternatively, in this embodiment of the present application, the second interaction stage may be referred to as a trust state.

[0206] The trust state in this embodiment of the present application can be understood as one in which the interaction relationship between the host vehicle and the conflicting object is clear, i.e., a judgment solution of the interaction between the host vehicle and the conflicting object is determined.

[0207] For example, in a competitive process between the self-vehicle and the competing target, as the competitive process continues, the self-vehicle approaches a judgment solution. When the difference between the judgment benefit of the approaching judgment solution and the judgment benefit of another judgment solution exceeds the threshold benefit difference, the competitive relationship is considered clear, and the transition from the crisis state to the trust state is completed.

[0208] Based on the contents shown in FIG. 12, further detailed explanation will be given to better explain the intelligent driving method provided in the present application, which is shown in FIG.

[0209] S1300: Recognition information is acquired.

[0210] S1301: Conflict targets are determined based on recognition information.

[0211] For details regarding the contents of step S1301, please refer to the descriptions of Figures 6 and 7. For the sake of brevity, the details will not be described again in this disclosure.

[0212] S1302: Obtain a set of judgment solutions for the host vehicle based on the recognition information.

[0213] S1303: Divide the judgment solution set into at least one judgment solution subset.

[0214] Apart from the above, each decision solution may include an action pair, which may include vehicle speed, vehicle offset direction, and so on.

[0215] Separately from the above, a judgment indicator of the judgment solution is determined based on the action pairs included in the judgment solution, and judgment solutions including the same judgment indicator in the judgment solution set are grouped into a judgment solution subset.

[0216] For example, a decision indicator is marked for each decision solution in the decision solution set as shown in Figure 14. For example, in the example scenario shown in Figure 14, after a decision indicator is marked for each decision solution in the decision solution set, the decision indicators of the decision solution set include: yield without avoidance, hurry through without avoidance, hurry through with avoidance, and yield with avoidance.

[0217] Then, the judgment solutions with the same decision indicator are grouped into one judgment solution subset. In the example scenario of Fig. 14, the judgment solution set is divided into four judgment solution subsets, namely, the judgment solution subset whose decision indicator is yield without avoidance, the judgment solution subset whose decision indicator is rush through without avoidance, the judgment solution subset whose decision indicator is rush through with avoidance, and the judgment solution subset whose decision indicator is yield with avoidance.

[0218] S1304: Determine an optimal judgment solution for each judgment solution subset. The optimal judgment solution is the judgment solution with the maximum judgment profit in the corresponding judgment solution subset.

[0219] For specific details of step S1304, please refer to the description of the contents of Figures 9 and 10. For the sake of brevity, details will not be described again in this disclosure.

[0220] S1305: Determine whether the ego-vehicle and the competing object satisfy the condition for entering the first interaction stage. If yes, execute S1306; if no, execute S1307.

[0221] For the specific content of step S1306, please refer to the above content description of the decision state machine module, and for the sake of brevity, the details will not be described again in this disclosure.

[0222] S1306: Obtain a first judgment solution used in a first interaction stage between the ego-vehicle and the conflict target, and continue to perform S1308.

[0223] A first decision solution to be used in a first interaction stage between the ego vehicle and the competing object is determined based on a plurality of optimal decision solutions included in a first hierarchy.

[0224] Since the judgment benefit of the optimal judgment solution in the first hierarchy is not less than the threshold benefit, the first judgment solution is selected from the optimal judgment solutions included in the first hierarchy, so that the selected first judgment solution may be more applicable at the current time, ensuring the safety and comfort of the interaction between the vehicle and the competing object.

[0225] Furthermore, in the first interaction stage, the acceleration of the first decision solution may further be planned to ensure that the ability to jump between various decision solutions may be available in future periods.

[0226] For example, when the ego vehicle is rushing past for a contender, rushing past with minimum acceleration can be implemented to ensure the ability to transition from rushing past to yielding. When the ego vehicle is yielding to a contender, yielding with maximum acceleration can be implemented to ensure the ability to transition from yielding to rushing past.

[0227] For example, as shown in FIG. 15, in the first interaction stage, it can be seen that the ST curve of the planned acceleration falls within the almond-shaped region of Sp and tp, i.e., the planned acceleration curve of the first judgment solution falls within the almond-shaped region of Sp and tp, and the planned acceleration includes the hurry through policy and the yielding policy.

[0228] Point SP shown in Fig. 15 can be understood as the final departure distance point and is used to indicate that the second interaction stage needs to be entered after point SP. Point TP shown in Fig. 15 can be understood as the final departure time point and is used to indicate that the second interaction stage needs to be entered after point TP. Focus P can be understood as the crisis departure point and is used to indicate the departure point from the first interaction stage to the second interaction stage. Line PG represents the planned route of the rush pass and line PY represents the planned route of the yield.

[0229] S1308: Perform a first interaction step with the conflict using the first decision solution, continually tracking the status of the conflict, and continue performing S1309.

[0230] S1309: Determine whether the ego-vehicle and the competing object satisfy the conditions for entering the second interaction stage, and if yes, execute S1307, and if no, return to S1308.

[0231] S1307: Obtain a second judgment solution to be used in a second interaction stage between the ego-vehicle and the conflict target, and continue to perform S1310.

[0232] In some embodiments, the second judgment solution is determined from a plurality of optimal judgment solutions included in the first tier.

[0233] As an example, assume that the threshold profit difference is 0.03, and the first hierarchy includes the optimal judgment solution 1 and the optimal judgment solution 2. In the initial stage of the interaction between the vehicle and the competing target, it is determined that the judgment profit of the target judgment solution 1 is 0.8, and the judgment profit of the target judgment solution 2 is 0.78. In this case, the difference between the judgment profit of the target judgment solution 1 and the judgment profit of the target judgment solution 2 is less than the threshold profit difference of 0.03. In this case, it cannot be determined which judgment solution is more suitable for this competition, and needs to be put into the first interaction stage. Assume that the optimal judgment solution 1 is determined as the first judgment solution of the first interaction stage.

[0234] As the interaction competition progresses, if it is determined that the judgment profit of the target judgment solution of the first tier is significantly greater than the judgment profit of another target judgment solution, it will enter the second interaction stage. For example, at a given time, the judgment profit of the optimal judgment solution 1 is 0.78, the judgment profit of the optimal judgment solution 2 is 0.82, and the judgment profit of the optimal judgment solution 2 exceeds the judgment profit of the optimal judgment solution 1 by 0.04, and the difference exceeds the threshold profit difference of 0.03. In this case, it may be determined that the judgment profit of the optimal judgment solution 2 of the first tier is currently significantly greater than the judgment profit of the optimal judgment solution 1, and the electronic device may determine that the optimal judgment solution 2 is more suitable for competition, enter the second interaction stage, and select the target judgment solution 2 as the second judgment solution for running.

[0235] As mentioned above, in this embodiment of the present application, there are multiple reasons for jumping from the first interaction stage to the second interaction stage, which may include in particular the following cases:

[0236] Assume that the current scene is the scene shown in FIG. 16, where the host vehicle is going straight and vehicles are about to merge at an intersection.

[0237] According to the above method, it is assumed that the first hierarchy includes a judgment solution subset of hasty passing with avoidance and a judgment solution subset of yielding without avoidance. The target judgment solution of hasty passing with avoidance can ensure high passability, and the target judgment solution of yielding without avoidance can ensure high safety. The difference between the judgment benefit of the optimal judgment solution 1 of hasty passing with avoidance and the judgment benefit of the optimal judgment solution 2 of yielding without avoidance is less than a threshold benefit difference.

[0238] In the first interaction phase, the judgment solution selected from the first hierarchy and used as the first judgment solution is hurry pass with avoidance (i.e., optimal judgment solution 1, which represents hurry pass with avoidance, is selected). In the process of the first interaction phase with the conflict object, the status of the conflict object is continuously tracked. If it is determined that the tracking result of the conflict object is not consistent with that predicted using the first judgment solution, a judgment solution suitable for the current status of the conflict object is obtained from the judgment solution set (e.g., yield without avoidance is selected), and the interaction jumps from the first interaction phase to the second interaction phase.

[0239] In some embodiments, a jump is performed from a first decision solution of a first interaction stage to a second decision solution of a second interaction stage, where the first decision solution is the same as the second decision solution.

[0240] Suppose the current scene is the scene shown in Fig. 17, where the ego vehicle is going straight and the vehicles are about to merge at an intersection. According to the above-mentioned method, suppose the first layer includes an optimal judgment solution 1 of hurrying through while avoiding and an optimal judgment solution 2 of yielding without avoiding. The optimal judgment solution 1 of hurrying through while avoiding can ensure high passing probability, and the optimal judgment solution 2 of yielding without avoiding can ensure high safety. The difference between the judgment benefit of the optimal judgment solution 1 of hurrying through while avoiding and the judgment benefit of the optimal judgment solution 2 of yielding without avoiding is less than a threshold benefit difference.

[0241] In the first interaction phase, the judgment solution selected from the first hierarchy and used as the first judgment solution is rush pass with avoidance (i.e., optimal judgment solution 1 representing rush pass with avoidance is selected). In the process of performing the first interaction phase with the conflicting object, the status of the conflicting object is continuously tracked. If it is determined that the tracking result of the conflicting object is always consistent with the first judgment solution, it is determined that the interaction is to jump from the first interaction phase to the second interaction phase, and the first hierarchy judgment solution of rush pass with avoidance continues to be used as the second judgment solution for traveling until the interaction is completed.

[0242] In some embodiments, if the benefit of the interaction policy being used exceeds the benefit of another decision solution in the decision solution set by a specified threshold in the first interaction phase (e.g., more than 10% of the benefit of the current interaction policy), a state jump may be initiated, i.e., the interaction jumps from the first interaction phase to the second interaction phase.

[0243] S1310: The electronic device performs a second interaction step with the conflicting object using the second decision solution until the interaction is completed.

[0244] According to the procedure shown in FIG. 13, the intelligent driving method provided in the embodiment of the present application may be implemented using any one of the electronic devices shown in FIG. 2 to FIG. 5.

[0245] Regarding the embodiment of the steps performed by the electronic device shown in FIG. 13, it is understood that according to the intelligent driving method provided in the embodiment of the present application, some steps shown in FIG. 13 may be omitted, some steps in FIG. 12 may be replaced with other steps, or the electronic device may further perform some steps not shown in FIG. 13.

[0246] Based on the same concept as above, the present application further provides an electronic device configured to perform the functions of the electronic device of the intelligent driving method described in the above method embodiment. Thus, the electronic device achieves the beneficial effects of the above method embodiment. The electronic device may include any of the structures shown in Figures 2 to 5, or may be implemented by a combination of any of the structures shown in Figures 2 to 5.

[0247] The electronic device shown in Fig. 2 may be a terminal or a vehicle, or may be a chip in a terminal or a chip in a vehicle. The electronic device can implement the intelligent driving method shown in Fig. 12 or Fig. 13 and the above-mentioned optional embodiments. The electronic device may include a processing module 210 and a communication module 220.

[0248] The processing module 210 may be configured to perform any one of steps S1200 to S1204 of the method shown in FIG. 11 and S1300 to S1310 of the method shown in FIG. 13, or any one of steps such as determining whether to enter a first interaction stage, determining whether to enter a second interaction stage, and determining a decision solution in the above-mentioned optional embodiment. The collection module 230 is configured to collect recognition information. For example, the collection module 230 may be configured to perform S1300 of the method shown in FIG. 13, or any step related to obtaining recognition information in the above-mentioned optional embodiment. For details, please refer to the detailed description of the example method. Details will not be described again in this disclosure.

[0249] The processing module 210 is configured to obtain a judgment solution set for the host vehicle based on the recognition information, obtain from the judgment solution set a first judgment solution to be used in a first interaction stage between the host vehicle and the competing object, detect that the host vehicle and the competing object satisfy a condition for entering a second interaction stage, obtain from the judgment solution set a second judgment solution to be used in a second interaction stage between the host vehicle and the competing object, and control the host vehicle to run based on the second judgment solution in the process of controlling the host vehicle to run based on the first judgment solution.

[0250] It is understood that the electronic device of the embodiment of the present application may be realized by software, for example, by a computer program having the above-mentioned functions or by instructions. The corresponding computer program or instructions may be stored in the memory of the terminal. The processor reads the corresponding computer program or instructions in the memory to realize the above-mentioned functions of the processing module 210 and / or the communication module 220 and / or the collection module 230. Alternatively, the electronic device of the embodiment of the present application may be realized by hardware. The processing module 210 may be a processor (for example, a CPU or a processor of a system chip), and the collection module 230 may include one or more of a camera device and a recognition device.

[0251] In a method that can be appropriately selected, the first judgment solution is a judgment solution used when the interaction relationship between the host vehicle and the competing object is unclear, and the second judgment solution is a judgment solution used when the interaction relationship between the host vehicle and the competing object is clear.

[0252] In an optional manner, the processing module 210 is further configured to continuously track the status of the conflict object in the process of the first interaction phase between the ego-vehicle and the conflict object.

[0253] Optionally, the conditions for entering the second interaction stage include: The duration of the first interaction phase exceeds a threshold duration, the decision profit difference between the decision solution with the first largest decision profit and the decision solution with the second largest decision profit in the decision solution set is not less than a threshold profit difference, the actual movement path of the competing object at a given time point is not consistent with the first decision solution, the actual movement path of the competing object at a given time point is consistent with the first decision solution, or the distance between the vehicle and the competing object is less than a threshold distance. Includes at least one of the following.

[0254] In an optional manner, the processing module 210: At least one object is determined based on the recognition information, a future movement path of the at least one object is acquired, and an object whose future movement path intersects with the future movement path of the host vehicle is determined as a competing object among the at least one object. The method is further configured as follows.

[0255] In an optional manner, the processing module 210: A left-right position offset sampling range and a front-back acceleration sampling range of the host vehicle and the competing object are obtained based on the recognition information, and a judgment solution set is obtained based on the left-right position offset sampling range and the front-back acceleration sampling range. The device is specifically configured as follows.

[0256] In an optional manner, the processing module 210: obtain a left / right position offset sampling range of the host vehicle and the competing object based on the recognition information and a first factor, obtain a forward / backward acceleration sampling range of the host vehicle and the competing object based on the recognition information and a second factor, the first factor including at least one of a road boundary and a stationary obstacle of an environment in which the host vehicle is located, and the second factor including at least one of a road speed limit, a jerk, and a competing object type of the environment in which the host vehicle is located; The device is specifically configured as follows.

[0257] In an optional manner, the processing module 210: Obtain at least one judgment solution subset by dividing the judgment solution set, each judgment solution subset including at least one judgment solution, determine a target judgment solution corresponding to each judgment solution subset, the target judgment solution being a judgment solution with a maximum judgment profit in the corresponding judgment solution subset, and obtain a first judgment solution from the at least one target judgment solution to be used in a first interaction stage between the ego-vehicle and the competing target; The device is specifically configured as follows.

[0258] In an optional manner, the processing module 210: Classify the subject judgment solutions whose judgment benefit is not less than the threshold benefit of at least one of the subject judgment solutions into a first tier, and classify the subject judgment solutions whose judgment benefit is less than the threshold benefit of at least one of the subject judgment solutions into a second tier; A first judgment solution to be used in a first interaction stage between the host vehicle and the competing object is determined from the object judgment solutions included in the first hierarchy. The device is specifically configured as follows.

[0259] In an optional manner, the processing module 210: Obtain a second judgment solution to be used in a second interaction stage between the vehicle and the competing object from the object judgment solutions included in the first hierarchy. The method is further configured as follows.

[0260] In an optional manner, the processing module 210: Determine a judgment benefit corresponding to each judgment solution in the judgment solution subset according to a benefit evaluation scheme, and determine a judgment solution with a maximum judgment benefit in the judgment solution subset as a target judgment solution corresponding to the judgment solution subset, where the benefit evaluation scheme includes a magnitude of at least one of safety, comfort, passability, right-of-way, danger area, and offset of the judgment solution; The device is specifically configured as follows.

[0261] In an optional manner, the benefit assessment method further includes a decision outcome history.

[0262] In a selective manner, each judgment solution includes one action pair group, and the action pair group includes vehicle acceleration and vehicle offset direction.

[0263] In an optional manner, the processing module 210: Determine a benefit evaluation method to be used for the interaction based on the correspondence between the driving environment and the benefit evaluation method, and determine a judgment benefit corresponding to each judgment solution in the judgment solution subset according to the benefit evaluation method to be used for the interaction. The device is specifically configured as follows.

[0264] In a possible design, the processing module 210 is further configured to determine that the ego-vehicle and the competing target satisfy conditions for entering a first interaction stage, the conditions for entering the first interaction stage including a judgment benefit difference between any two target judgment solutions in the first tier being less than a threshold benefit difference.

[0265] In a possible design, the processing module 210 is further configured to decide to perform the second stage if the difference between the judgment benefit of the target judgment solution in the first tier and the judgment benefit of each remaining target judgment solution in the first tier exceeds a threshold benefit difference.

[0266] In a possible design, the processing module 210 is further configured to adjust the decision solution used in the first interaction phase if it determines that the tracking result of the competing object at a predetermined time point in the first interaction phase is inconsistent with the movement path at the predetermined time point predicted using the first decision solution.

[0267] In a possible design, the first decision solution is the same as the second decision solution.

[0268] It is understood that for details of the processing of the electronic device in this embodiment of the present application, please refer to Figures 12 and 13 and the related description of the method embodiment of the present application, and the details will not be described again in this disclosure.

[0269] The electronic device shown in Fig. 4 may be a vehicle. The vehicle control device may implement the intelligent driving method shown in Fig. 12 or Fig. 13 and the optional embodiments described above. The electronic device may include at least one of a processor, a memory, an interface circuit, or a human-computer interaction device. It should be understood that Fig. 4 shows only one processor, only one memory, only one interface circuit, and only one (or only one type) of collection device. The electronic device may include other numbers of processors and other numbers of interface circuits.

[0270] The interface circuit is configured to connect the electronic device to another component of the terminal or vehicle, such as a memory, another processor, or a projection device. The processor is configured to perform signal interaction with the other component through the interface circuit. The interface circuit may be an input / output interface of the processor.

[0271] For example, the processor reads a computer program or instruction in a memory coupled to the processor through an interface circuit, and decodes and executes the computer program or instruction. It is understood that the computer program or instruction may include the aforementioned functional program or may include the aforementioned functional program of the electronic device. When the corresponding functional program is decoded and executed by the processor, the electronic device can be enabled to implement the solution in the intelligent driving method provided in the embodiment of the present application.

[0272] Alternatively, the functional program may be stored in a memory external to the electronic device, in which case the electronic device may not include a memory, and the memory temporarily stores some or all of the contents of the functional program when the functional program is decoded and executed by the processor.

[0273] Alternatively, the function program is stored in a memory in the electronic device. When the memory of the electronic device stores the aforementioned function program, the electronic device may be disposed in the electronic device of the embodiment of the present application.

[0274] Alternatively, the function program may be stored in a memory external to the road condition television apparatus, with other parts of the function program being stored in a memory within the electronic device.

[0275] It should be understood that a processor may be a chip, for example, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), a central processor unit (CPU), a network processor (NP), a digital signal processor (DSP), a micro controller unit (MCU), a programmable logic device (PLD), or another integrated chip.

[0276] It should be noted that the processor of the embodiments of the present application may be an integrated circuit chip and has signal processing capabilities. In one implementation process, the steps of the above-mentioned method embodiments may be implemented using hardware integrated logic circuits of the processor, or may be implemented using instructions in the form of software. The processor may be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component. The processing module may implement or perform the methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor, the processor may be any conventional processor, and so on. The steps of the methods disclosed with reference to the embodiments of the present application may be directly implemented and realized using a hardware decoding processor, or may be implemented and realized using a combination of hardware and software modules of the decoding processor. The software modules may be located in a storage medium that is a mature technology in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, or a register. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with the processor's hardware.

[0277] It is understood that the memory of the present embodiment of the present application may be volatile or non-volatile memory, or may include volatile and non-volatile memory. The non-volatile memory may be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory may be random access memory (RAM) used as an external cache. Throughout the description of the examples (but without imposing limitations), many forms of RAM may be used, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus dynamic random access memory (DR RAM). It should be noted that the memory of the systems and methods described herein includes, but is not limited to, the above memories and any other suitable type of memory.

[0278] It should be understood that when the electronic device is implemented using the structure shown in FIG. 4, the memory may store computer programs or instructions, and the processor executes the computer programs or instructions stored in the memory to perform the operations performed by the processing module 210 when the electronic device is implemented using the structure shown in FIG. 2. When the electronic device is implemented using the structure shown in FIG. 2, the collection device may perform the operation of collecting recognition information. Alternatively, the processor and memory shown in FIG. 4 may be used to implement the processing module 210 shown in FIG. 2, or the processing module 210 shown in FIG. 2 may include a processor and a memory, or the processor executes the computer programs or instructions stored in the memory to perform the operations performed by the processing module 210 shown in FIG. 2, and / or the collection device shown in FIG. 4 may implement the collection module 230 shown in FIG. 2, or the processing module 210 shown in FIG. 2 may include the collection device shown in FIG. 4, or the collection device performs the operations performed by the collection module 230 shown in FIG. 2.

[0279] When the electronic device is implemented using the structure shown in Figure 5, one or more of the competing obstacle screening module, the competing object status tracking module, the conflict determination module, and the motion planning module may perform operations performed by the processing module 210 when the electronic device is implemented using the structure shown in Figure 2. When the electronic device is implemented using the structure shown in Figure 5, the operations performed by the competing obstacle screening module, the competing object status tracking module, the conflict determination module, and the motion planning module, respectively, are referred to the description of the procedures shown in Figure 12 and / or Figure 13. Details will not be described again in this disclosure.

[0280] It is understood that the structures of the electronic devices shown in Figures 2 to 5 may be combined with each other. The relevant design details of the electronic devices shown in Figures 2 to 5 and the optional embodiment are referred to each other, and the relevant design details of the intelligent driving method shown in any one of Figures 2 to 5 and the optional embodiment are referred to. The details will not be repeated.

[0281] Based on the same concept as above, the present application provides a computing device including a processor, the processor being connected to a memory, the memory being configured to store computer programs or instructions, the processor being configured to execute the computer programs stored in the memory, such that the computing device performs the method of the above method embodiment.

[0282] Based on the same concept as above, the present application provides a computer-readable storage medium, which stores a computer program or instruction, which, when executed, enables a computing device to perform the method of the above method embodiment.

[0283] Based on the same concept as above, the present application provides a computer program product, which, when executed by a computer, enables the computing device to perform the method of the above method embodiments.

[0284] Based on the same concept as above, the present application provides a chip, which is connected to a memory and configured to read and execute computer programs or instructions stored in the memory, so that the computing device performs the method of the above method embodiment.

[0285] Based on the same concept as above, an embodiment of the present application provides an apparatus, which includes a processor and an interface circuit, the interface circuit being configured to receive a computer program or instruction and pass the computer program or instruction to the processor, which runs the computer program or instruction to perform the method of the above method embodiment.

[0286] Based on the same concept as the above, an embodiment of the present application provides a vehicle, which includes: a recognition module configured to acquire recognition information; and a processing module configured to acquire a judgment solution set of the host vehicle based on the recognition information, acquire a first judgment solution from the judgment solution set for a first interaction stage between the host vehicle and a conflicting object, and control the host vehicle to run based on the first judgment solution; detect that the host vehicle and the conflicting object satisfy a condition for entering a second interaction stage, acquire a second judgment solution from the judgment solution set for a second interaction stage between the host vehicle and the conflicting object, and control the host vehicle to run based on the second judgment solution.

[0287] Apart from the above, in the embodiments of the present application, the division into modules is used as an example, which is merely a logical functional division. In actual implementation, other division methods may be used. In addition to the above, the functional modules of the embodiments of the present application may be integrated into one processor, each of the modules may exist physically alone, or two or more modules may be integrated into one module. The integrated modules may be realized in the form of hardware or in the form of software functional modules.

[0288] It should be appreciated by those skilled in the art that the embodiments of the present application may be provided as a method, a system, or a computer program product. Thus, the present application may employ hardware-only embodiments, software-only embodiments, or embodiments that combine software and hardware. In addition to the above, the present application may employ the form of a computer program product embodied in one or more computer-usable storage media (including, but not limited to, disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0289] The present application has been described with reference to flowcharts and / or block diagrams of the methods, devices (systems) and computer program products of the present application. It will be understood that computer program instructions may be used to implement each process and / or each block of the flowcharts and / or block diagrams, and combinations of the processes and / or blocks of the flowcharts and / or block diagrams. The computer program instructions may be provided for a processor of a general purpose computer, a special purpose computer, an embedded processor, or any other programmable data processing device to generate a machine, whereby the instructions executed by the processor of the computer or any other programmable data processing device generate an apparatus that implements the specified functionality of one or more processes in the flowcharts and / or one or more blocks in the block diagrams.

[0290] The computer program instructions may be stored in a computer readable memory that can instruct a computer or any other programmable data processing device to act in a particular manner, such that the instructions stored in the computer readable memory produce an artifact that includes an instruction apparatus that implements a particular function in one or more processes in the flowcharts and / or one or more blocks in the block diagrams.

[0291] Alternatively, computer program instructions may be loaded into a computer or other programmable data processing device such that a sequence of operations and steps are performed by the computer or other programmable device to generate a computer-implemented process. Thus, the instructions that execute on the computer or other programmable device provide steps for implementing a particular function of one or more procedures in one or more blocks in the flowcharts and / or block diagrams.

[0292] It is obvious that those skilled in the art can make various modifications and variations to this application without departing from the scope of this application. This application intends to cover these modifications and variations of this application as long as they fall within the scope of protection defined by the following claims and their equivalent technologies. [Explanation of symbols]

[0293] 210 Processing Module 220 Communication Module 230 Collection Module 240 Execution Module 250 Memory Module 260 display screen

Claims

1. 1. An intelligent driving method, comprising: obtaining a set of judgment solutions for the vehicle based on the recognition information; obtaining a first judgment solution from the judgment solution set, the first judgment solution being used in a first interaction stage between the ego vehicle and a competing target; determining that a condition for the ego-vehicle and the competing target to enter a second interaction phase is met in the process of controlling the ego-vehicle to travel based on the first judgment solution; obtaining a second judgment solution from the judgment solution set, the second judgment solution being used in the second interaction phase between the ego vehicle and the conflict target; controlling the host vehicle to travel based on the second judgment solution; A method for providing the above.

2. The method according to claim 1 , wherein the first interaction stage is a stage in which an interaction relationship between the host vehicle and the competing object is unclear, and the second interaction stage is a stage in which an interaction relationship between the host vehicle and the competing object is clear.

3. The method comprises: Continually tracking the status of the conflicting object during the process of the first interaction phase between the ego vehicle and the conflicting object. The method of claim 1 or 2, further comprising:

4. The condition for entering the second interaction phase is: a duration of the first interaction phase exceeding a threshold duration; a judgment profit difference between the judgment solution with the maximum judgment profit in the judgment solution set and any other judgment solution in the judgment solution set is not less than a threshold profit difference; the actual path of movement of the conflicting object at a given time is inconsistent with the first determination solution; the actual path of movement of the conflicting object at a given time is consistent with the first determination solution; or The distance between the host vehicle and the competing object is less than a threshold distance. The method according to any one of claims 1 to 3, comprising one of the following steps:

5. The method comprises: determining at least one object based on the recognition information; obtaining a future movement path of the at least one object; acquiring a future moving path of the host vehicle; determining, as the competing object, an object whose future movement path intersects with the future movement path of the host vehicle among the at least one object; 5. The method of claim 1, further comprising:

6. The step of obtaining a set of judgment solutions for the vehicle based on the recognition information includes: acquiring a left / right direction position offset sampling range and a front / rear direction acceleration sampling range of the host vehicle and the competing object based on the recognition information; obtaining the judgment solution set based on the lateral position offset sampling range and the longitudinal acceleration sampling range; 6. The method according to claim 1 , comprising:

7. the step of acquiring a left / right direction position offset sampling range and a front / rear direction acceleration sampling range of the host vehicle and the competing object based on the recognition information, acquiring the lateral direction position offset sampling range of the host vehicle and the competing object based on the recognition information and a first factor; acquiring the longitudinal acceleration sampling range of the host vehicle and the competing object based on the recognition information and a second factor; Equipped with 7. The method of claim 6, wherein the first factor comprises at least one of road boundaries and stationary obstacles of an environment in which the host vehicle is located, and the second factor comprises at least one of road speed limits and jerk of the environment in which the host vehicle is located.

8. The step of obtaining a first judgment solution from the judgment solution set to be used in a first interaction stage between the host vehicle and a competing target includes: obtaining at least one judgment solution subset by dividing the judgment solution set, each judgment solution subset comprising at least one judgment solution; determining an optimal judgment solution for each judgment solution subset, the optimal judgment solution being the judgment solution with the maximum judgment profit in the corresponding judgment solution subset; obtaining the first judgment solution from at least one optimal judgment solution for use in the first interaction phase between the ego vehicle and the competing object; 8. The method according to claim 1 , comprising:

9. The step of obtaining the first judgment solution to be used in the first interaction phase between the ego vehicle and the competing target from at least one optimal judgment solution includes: classifying optimal judgment solutions whose judgment profit is not less than a threshold profit of the at least one optimal judgment solution into a first tier and classifying optimal judgment solutions whose judgment profit is less than the threshold profit of the at least one optimal judgment solution into a second tier; determining the first judgment solution to be used in the first interaction stage between the ego vehicle and the competing object from the optimal judgment solutions provided in the first hierarchy; The method of claim 8 comprising:

10. The method comprises: A step of obtaining the second judgment solution to be used in the second interaction stage between the host vehicle and the competing object from the optimal judgment solution provided in the first hierarchy. The method of claim 9 , further comprising:

11. The step of determining an optimal decision solution corresponding to each decision solution subset comprises: determining a judgment benefit corresponding to each judgment solution in the judgment solution subset according to a benefit evaluation scheme; determining a judgment solution having a maximum judgment profit in the judgment solution subset as the optimal judgment solution corresponding to the judgment solution subset; Equipped with The method according to any one of claims 8 to 10, wherein the benefit assessment scheme comprises a magnitude of at least one of safety, comfort, passability, right-of-way, danger area and offset of the determined solution.

12. The method of claim 11 , wherein the benefit assessment scheme further includes a decision outcome history.

13. The method of claim 1 , wherein each decision solution comprises an action pair group.

14. 14. A computer readable storage medium storing a computer program or instructions which, when executed by a computing device, enable the computing device to perform the method of any one of claims 1 to 13.

15. A chip comprising at least one processor and an interface; the interface is configured to provide computer programs, instructions, or data to the at least one processor; A chip, wherein the at least one processor is configured to execute the computer program or instructions, thereby performing the method of any one of claims 1 to 13.

16. A vehicle, a recognition module configured to obtain recognition information; a processing module configured to: obtain a judgment solution set for the host vehicle based on the recognition information; obtain from the judgment solution set a first judgment solution to be used in a first interaction stage between the host vehicle and a conflicting object; and, in a process of controlling the host vehicle to travel based on the first judgment solution, detect that the host vehicle and the conflicting object satisfy a condition for entering a second interaction stage; obtain from the judgment solution set a second judgment solution to be used in a second interaction stage between the host vehicle and the conflicting object; and control the host vehicle to travel based on the second judgment solution; A vehicle equipped with.

Images