Vehicle driving control method, device, equipment and medium

Abstract

The invention discloses a vehicle driving control method, device and equipment and a medium, and belongs to the technical field of vehicles. The method comprises the following steps: acquiring a first prediction trajectory set corresponding to a plurality of blind area targets in a vehicle blind area; for each blind area target, judging whether the blind area target and the vehicle have a collision risk or not based on a first prediction trajectory set corresponding to the blind area target and a target planning trajectory of the vehicle; the target planning trajectory is a trajectory in a plurality of candidate planning trajectories along the current driving direction of the vehicle, and the plurality of candidate planning trajectories are determined based on the state information of the vehicle; based on the judgment result, determining a target trajectory set corresponding to the risk blind area target with the collision risk from the first prediction trajectory set; generating collision information based on the target trajectory set and the plurality of candidate planning trajectories of the vehicle; a first control instruction is generated based on the collision information, and the driving state of the vehicle is controlled based on the first control instruction.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a vehicle driving control method, device, equipment and medium. Background Technology

[0002] During vehicle operation, there are many areas around the vehicle where the view is obstructed due to the structure of buildings. These areas are called blind spots. Other vehicles, pedestrians, and other targets may be present in the blind spots. If not careful, these targets may collide with the vehicle, causing personal injury and property damage.

[0003] Therefore, it is necessary to provide a solution that can prevent vehicles from colliding with blind spot targets in scenarios where there is a potential risk of collision, thereby improving driving safety. Summary of the Invention

[0004] In view of the above-mentioned problem of how to prevent vehicles from colliding with blind spot targets in order to improve driving safety, this application is made to provide a vehicle driving control method, device, equipment and medium that can prevent vehicles from colliding with blind spot targets in order to improve driving safety.

[0005] In a first aspect, this application provides a vehicle driving control method, the method comprising: Obtain a first set of predicted trajectories corresponding to multiple blind spot targets within the vehicle's blind spot; the first set of predicted trajectories includes at least one predicted trajectory. For each blind spot target, based on the first predicted trajectory set corresponding to the blind spot target and the target planned trajectory of the vehicle, it is determined whether there is a collision risk between the blind spot target and the vehicle; the target planned trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planned trajectories, and the multiple candidate planned trajectories are determined based on the state information of the vehicle; Based on the judgment results, the target trajectory sets corresponding to the risk blind zone targets with collision risk are determined from the first predicted trajectory set; Based on the target trajectory set and multiple candidate planned trajectories of the vehicle, collision information is generated; A first control command is generated based on the collision information, and the driving state of the vehicle is controlled based on the first control command.

[0006] In one embodiment, generating collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle includes: Construct an environment model; the environment model includes attribute information of various environmental targets in the current environment in which the vehicle is located, which pose a collision risk to the vehicle, and the attribute information includes a first identifier and state information; Associate each target trajectory set with the corresponding environmental target in the environmental model; The target trajectory set associated with each environmental target is respectively used as the second predicted trajectory set of the environmental target; Collision information is generated based on the second predicted trajectory set and the candidate planned trajectory.

[0007] In one embodiment, the construction environment model includes: For each risk blind zone target, determine the estimated collision points between the target trajectory set corresponding to the risk blind zone target and the target planned trajectory; Based on the state information of the estimated collision point, the state information of the environmental target is generated; and based on the second identifier of the risk blind zone target, the first identifier of the environmental target is generated. The environmental model is constructed based on the state information of each environmental target and the first identifier.

[0008] In one embodiment, associating each set of target trajectories with the corresponding environmental targets in the environment model includes: For each environmental target in the environmental model, a matching target identifier is determined from a plurality of second identifiers based on the first identifier of the environmental target; Determine the first predicted trajectory set of the blind zone target indicated by the target identifier; Associate the first predicted trajectory set of the blind zone target indicated by the target identifier with the environmental target.

[0009] In one embodiment, generating state information of the environmental target based on the estimated collision point state information includes: For each target in the risk blind zone, the current predicted state information of the target in the risk blind zone is determined based on the state information of the predicted collision point; Based on the predicted state information, the state information of the environmental target is generated.

[0010] In one embodiment, the collision information is the duration between the current time and the time corresponding to the vehicle collision. Generating the collision information based on the second predicted trajectory set and the candidate planned trajectory includes: For each environmental target, based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle, it is determined whether there is a collision risk between the environmental target and the vehicle; If there is a risk of collision, determine the current moment and the duration between the moment the vehicle collides with the environmental target; The collision information is generated based on the duration corresponding to each environmental target.

[0011] In one embodiment, determining the duration between the current moment and the moment of collision includes: Determine the target prediction trajectory in the second predicted trajectory set of the environmental target that collided, as well as the pre-collision planned trajectory among multiple candidate planned trajectories; Indeed, the actual collision point between the predicted target trajectory and the pre-collision planned trajectory; Based on the actual collision point, the target predicted trajectory, and the pre-collision planned trajectory, the duration between the current moment and the moment of collision is determined.

[0012] In one embodiment, generating the collision information based on the duration corresponding to each environmental target includes: The minimum duration among the durations corresponding to each environmental target is determined to obtain collision information.

[0013] In one embodiment, generating the first control command based on the collision information includes: If the minimum duration in the collision information is not greater than the first preset duration, a first driving control command is generated. The first driving control command includes a braking command or an emergency steering command. The emergency steering command is used to indicate that the steering angle is greater than the preset steering angle within a preset unit time. If the minimum duration in the collision information is greater than the first preset duration and less than the second preset duration, then based on the target predicted trajectory and the pre-collision planned trajectory, a final planned trajectory is determined, and a second driving control command is generated based on the final planned trajectory; the second driving control command includes a deceleration command and / or a smooth steering command, the smooth steering command being used to indicate that the steering angle is less than a preset steering angle within a preset unit time.

[0014] In one embodiment, after determining whether there is a collision risk between the environmental target and the vehicle based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle, the method further includes: If there is no risk of collision between any environmental target and the vehicle, the state information of each environmental target in the environmental model is visualized; and a second control command is generated based on the state information of each environmental target, and the driving state of the vehicle is controlled based on the second control command.

[0015] In one embodiment, generating the state information of the environmental target based on the predicted state information includes: Obtain target configuration parameters; the target configuration parameters are the configuration parameters of the state information of the real target collected by the sensors in the vehicle. The configuration parameters of the predicted state information are set to the target configuration parameters to obtain the state information of the environmental target; The target configuration parameters include: target format, transmission interface, and acquisition frequency.

[0016] In one embodiment, obtaining the first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot includes: Send an information acquisition request to the roadside unit so that the roadside unit can collect the current status information of each blind spot target among the multiple blind spot targets; Receive the current status information of each blind spot target fed back by the roadside unit; For each blind zone target, trajectory prediction is performed on the blind zone target based on the current state information of the blind zone target to obtain the corresponding first predicted trajectory set.

[0017] Secondly, this application provides a vehicle driving control device, the device comprising: The acquisition module is used to acquire a first set of predicted trajectories corresponding to multiple blind spot targets within the vehicle's blind spot; the first set of predicted trajectories includes at least one predicted trajectory. The judgment module is used to determine whether there is a collision risk between the blind spot target and the vehicle based on the first predicted trajectory set corresponding to the blind spot target and the target planned trajectory of the vehicle for each blind spot target; the target planned trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planned trajectories, and the multiple candidate planned trajectories are determined based on the state information of the vehicle; The determination module is used to determine, based on the judgment result, the target trajectory set corresponding to the risk blind zone targets with collision risk from the first predicted trajectory set; The generation module is used to generate collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle; The control module is used to generate a first control command based on the collision information, and to control the driving state of the vehicle based on the first control command.

[0018] Thirdly, this application provides an electronic device, including: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method as described in the first aspect.

[0019] Fourthly, this application provides a computer-readable storage medium storing computer instructions for causing the computer to perform the method described in the first aspect.

[0020] The technical solutions provided in this application embodiment have at least the following technical effects or advantages: This application obtains a first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot area. The first predicted trajectory set includes at least one predicted trajectory. For each blind spot target, based on the first predicted trajectory set corresponding to the blind spot target and the vehicle's target planning trajectory, it determines whether there is a collision risk between the blind spot target and the vehicle. The target planning trajectory is the trajectory along the vehicle's current driving direction from among multiple candidate planning trajectories. These multiple candidate planning trajectories are determined based on the vehicle's state information. Based on the judgment result, the target trajectory set corresponding to the risky blind spot targets can be determined from the first predicted trajectory set. This allows for the selection of target trajectory sets from the first predicted trajectory set, achieving trajectory set filtering. Based on the target trajectory set predicted for the risky blind spot targets and the vehicle's multiple candidate planning trajectories, collision information is generated. A first control command is generated based on the collision information, and the vehicle's driving state is controlled based on the first control command. In other words, based on the predicted trajectory obtained from the trajectory prediction of the blind spot target and the vehicle's candidate planning trajectory, control commands are generated to control the vehicle's driving state to prevent collisions with blind spot targets and improve driving safety. Meanwhile, this application enables the filtering of trajectory sets, reducing the number of trajectory sets that need to be processed when generating collision information. This allows for the rapid generation of control commands, which in turn enables the rapid control of the vehicle's driving status when there is a risk of collision, thereby improving driving safety.

[0021] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0022] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a flowchart of a vehicle driving control method provided in an embodiment of this application; Figure 2 This is a flowchart of another vehicle driving control method provided in the embodiments of this application; Figure 3 This is a flowchart of another vehicle driving control method provided in the embodiments of this application; Figure 4 This is a schematic diagram of the vehicle driving control device provided in the embodiments of this application; Figure 5 This is a schematic diagram of the electronic device. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the embodiments of this disclosure and the specific features in the embodiments are detailed descriptions of the technical solutions of this application, rather than limitations on the technical solutions of this application. Unless otherwise specified, the embodiments of this application and the technical features in the embodiments can be combined with each other.

[0024] Reference Figure 1 This application provides a vehicle driving control method, which is executed by an on-board unit in the vehicle, and the method specifically includes: Step S101: Obtain the first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot; the first predicted trajectory set includes at least one predicted trajectory. In this embodiment, there are static and dynamic targets within the blind zone. Static targets can be buildings, tree stumps, etc., within the blind zone, while dynamic targets can be vehicles, pedestrians, etc. In this application, "blind zone targets" refers specifically to dynamic targets within the blind zone.

[0025] Each blind zone target corresponds to a first predicted trajectory set, and any first predicted trajectory set includes at least one predicted trajectory.

[0026] Step S102: For each blind spot target, based on the first predicted trajectory set corresponding to the blind spot target and the vehicle's target planning trajectory, determine whether there is a collision risk between the blind spot target and the vehicle; the target planning trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planning trajectories, and the multiple candidate planning trajectories are determined based on the vehicle's state information; For each blind spot target, if any of the first predicted trajectories of the blind spot target intersects with the vehicle's target planned trajectory, then it is determined that there is a collision risk between the blind spot target and the vehicle.

[0027] Based on the vehicle's current state information, trajectory prediction can be performed, and multiple candidate trajectories can be predicted.

[0028] The vehicle's status information also includes the vehicle's current driving direction. The target planned trajectory is the trajectory along the current driving direction from among multiple candidate planned trajectories. That is, the target planned trajectory only considers the trajectory of the vehicle traveling straight along the current driving mode, and does not consider the trajectory of the vehicle turning. The multiple candidate planned trajectories specifically include: the trajectory of the vehicle traveling straight along the current driving mode, and the trajectory of the vehicle turning.

[0029] Step S103: Based on the judgment results, determine the target trajectory sets corresponding to the risk blind zone targets with collision risk from the first predicted trajectory set; For each blind spot target, if there is a collision risk between the blind spot target and the vehicle, the blind spot target is marked as a risky blind spot target. The first predicted trajectory set of each risky blind spot target is used as the target trajectory set.

[0030] Step S104: Generate collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle; For each risk blind zone target, based on the target trajectory set corresponding to the risk blind zone target and multiple candidate planned trajectories of the vehicle in real time, collision information corresponding to each risk blind zone target is generated; Step S105: Generate a first control command based on the collision information, and control the driving state of the vehicle based on the first control command.

[0031] This application can generate a first control command that the vehicle can understand, thereby controlling the driving state of the vehicle based on the first control command.

[0032] As can be seen, this application obtains a first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot area. The first predicted trajectory set includes at least one predicted trajectory. Based on the first predicted trajectory set corresponding to each blind spot target and the vehicle's target planning trajectory, it determines whether there is a collision risk between the blind spot target and the vehicle. The target planning trajectory is the trajectory along the vehicle's current driving direction from among multiple candidate planning trajectories. These multiple candidate planning trajectories are determined based on the vehicle's state information. Therefore, based on the judgment result, the target trajectory set corresponding to the risky blind spot targets with collision risk can be determined from the first predicted trajectory set. This allows for the filtering of the trajectory set. Based on the predicted target trajectory set of the risky blind spot targets and the vehicle's multiple candidate planning trajectories, collision information is generated. A first control command is generated based on the collision information, and the vehicle's driving state is controlled based on the first control command. In other words, based on the predicted trajectory obtained from the trajectory prediction of the blind spot targets and the vehicle's candidate planning trajectories, control commands are generated to control the vehicle's driving state to prevent collisions with blind spot targets and improve driving safety. Meanwhile, this application enables the filtering of trajectory sets, reducing the number of trajectory sets that need to be processed when generating collision information. This allows for the rapid generation of control commands, which in turn enables the rapid control of the vehicle's driving status when there is a risk of collision, thereby further improving driving safety.

[0033] Meanwhile, some current solutions utilize vehicle-to-everything (V2X) technology to enable vehicles to receive real-time status information of targets within blind spots from roadside units. When the real-time status information indicates a collision risk between the vehicle and a target within the blind spot, a warning is issued to the driver, alerting them to the blind spot risk and allowing the driver to take over vehicle control. However, in emergency situations, the driver may not be able to react in time to avoid a collision. Overall, this solution offers relatively poor vehicle driving safety.

[0034] The solution proposed in this application involves predicting the trajectory of a target in the blind spot, obtaining the predicted trajectory, and then combining it with the vehicle's candidate planned trajectory to generate collision information. This allows for the prediction of potential collision scenarios between the vehicle and the target in the blind spot before an actual collision occurs. Based on this collision information, a first control command that the vehicle can recognize is generated, and the vehicle's driving state is then controlled according to this first control command. This approach, on the one hand, allows time for the vehicle to control its driving state by anticipating potential collisions before they actually happen; on the other hand, by generating a first control command that the vehicle can recognize, the vehicle's driving state is controlled by the driver, avoiding situations where the driver's reaction time is insufficient when manually controlling the vehicle. Therefore, compared to existing solutions, this application improves vehicle driving safety.

[0035] Furthermore, in the embodiments of this application, after obtaining the target trajectory set, the identification and processing of the original, dispersed target trajectory set requires modification of the input interface and internal logic of the planning algorithm. This violates the original intention of "low intrusion and easy integration" and greatly increases the implementation cost. However, the solution of this application: by constructing an environmental model compatible with the vehicle unit's control system, the original target trajectory set is associated with the corresponding environmental targets in the environmental model, obtaining a predicted trajectory set for each environmental target in the environmental model. Thus, the vehicle unit can directly identify and process the predicted trajectory set of environmental targets without modifying the input interface and internal logic of the planning algorithm, achieving the goal of low intrusion and easy integration, and reducing the implementation cost. Specifically: In one embodiment, collision information is generated based on a set of target trajectories and multiple candidate planned trajectories of the vehicle, including: constructing an environment model; the environment model includes attribute information of various environmental targets in the current environment in which the vehicle is located, which pose a collision risk to the vehicle, and the attribute information includes a first identifier and state information; associating each set of target trajectories with the corresponding environmental targets in the environment model; using the set of target trajectories associated with each environmental target as a second predicted trajectory set for the environmental target; and generating collision information based on the second predicted trajectory set and the candidate planned trajectories.

[0036] The environmental model includes multiple environmental targets, each of which poses a collision risk with the vehicle. The attribute information of the environmental targets includes a first identifier and state information. The first identifier is used to uniquely identify an environmental target, and the state information includes real-time location, speed, and type. The type can be used to indicate whether the environmental target is a pedestrian or a vehicle.

[0037] In one embodiment, the construction of the environment model includes: for each risk blind zone target, determining the estimated collision points between the target trajectory set corresponding to the risk blind zone target and the target planned trajectory; generating state information of the environment target based on the state information of the estimated collision points; generating a first identifier of the environment target based on the second identifier of the risk blind zone target; and constructing the environment model based on the state information of each environment target and the first identifier.

[0038] In this embodiment, based on the target trajectory set of the target in the risk blind zone, a representative state representing its potential threat is determined, such as the estimated collision point on the predicted trajectory where a collision with the candidate planned trajectory of the vehicle may occur. If there are at least two predicted trajectories that collide with the target planned trajectory in the target trajectory set of the risk blind zone, then calculate the collision probability of each predicted trajectory with the target planned trajectory, and take the collision point of the predicted trajectory with the target planned trajectory with the higher collision probability as the estimated collision point. If there is only one collision point between the predicted trajectory and the planned trajectory of the target in the risk blind zone, then the collision point between the predicted trajectory and the planned trajectory of the target in the risk blind zone will be taken as the estimated collision point. After the roadside equipment assigns a unique second identifier to the blind spot target, a first identifier for the environmental target is generated based on the second identifier of the risk blind spot target. This first identifier includes: key identifier information based on the second identifier of the risk blind spot target, which indicates a unique target within the blind spot; and the first identifier generated based on the identifier representation of the environmental target and the key identifier information. The identifier representation of the environmental target differs from that of the risk blind spot target.

[0039] An environmental model is constructed based on the state information of the environmental targets corresponding to each risk blind zone target and their corresponding first identifier. Specifically, for environmental targets that do not exist in the initial environmental model, the state information of the environmental targets and their corresponding first identifiers are added to the initial environmental model. For environmental targets that exist in the initial environmental model, the original state information of the environmental targets in the initial environmental model is updated with the state information of the environmental targets.

[0040] Specifically, for each risk blind zone target, the status information of the corresponding environmental target and the first identifier constitute the attribute information of the environmental target in the environmental model.

[0041] In one embodiment, associating each target trajectory set with the corresponding environmental target in the environmental model includes: for each environmental target in the environmental model, determining a matching target identifier from multiple second identifiers based on a first identifier of the environmental target; determining a first predicted trajectory set of the blind zone target indicated by the target identifier; and associating the first predicted trajectory set of the blind zone target indicated by the target identifier with the environmental target.

[0042] Based on the first identifier of the environmental objective, a matching target identifier is determined from a plurality of second identifiers, including: if the key identifier information in any second identifier is the same as the key identifier information in the first identifier, then the second identifier is taken as the matching target identifier.

[0043] In this embodiment, the blind zone target indicated by the target identifier is the aforementioned risk blind zone target. The first predicted trajectory set of the blind zone target is the aforementioned target trajectory set.

[0044] As can be seen, this application can find the first predicted trajectory set corresponding to the first identifier from the pre-generated first predicted trajectory set based on the first identifier of the environmental target, and associate it with the environmental target. In this way, the predicted trajectory set of each environmental target in the environmental model can be obtained, namely the second predicted trajectory set. Based on the second predicted trajectory set and the candidate planned trajectory, it is possible to determine whether there is a collision risk between the vehicle and the environmental target and generate collision information.

[0045] As can be seen, this application obtains a second predicted trajectory set of environmental targets by associating the first predicted trajectory set (i.e., the original predicted trajectory set) of the blind zone targets indicated by the target identifier with the environmental targets. In this way, the original predicted trajectory set becomes a predicted trajectory set of each environmental target in the environmental model that the vehicle can directly identify and process, rather than the original predicted trajectory set that requires modification of the input interface and internal logic of the planning algorithm to identify and process.

[0046] In one embodiment, after obtaining the predicted collision points of the target trajectory set corresponding to the target blind zone target and the target planned trajectory, the state information of the starting point corresponding to the target trajectory set is also solved in reverse. This is because the position of the blind zone target perceived in real time is an "observation with error," not the true position. The reverse solution is to calibrate the starting point of the predicted trajectory set to a true and reasonable position that conforms to the collision constraints, so that the target trajectory set and collision calculation form a closed loop and match the physical laws. Specifically: The step of generating environmental target status information based on the predicted collision point status information includes: for each risk blind zone target, determining the current predicted status information of the risk blind zone target based on the predicted collision point status information; and generating the environmental target status information based on the predicted status information.

[0047] The predicted state information of the collision point includes: position, velocity, and type; the definition of type is similar to that above and will not be repeated here. For each risk blind zone target, based on the state information of the predicted collision point, the state information of the starting point corresponding to the target trajectory set is solved in reverse. The state information of this starting point is the predicted current state information of the risk blind zone target, that is, the predicted state information of the risk blind zone target. The predicted state information of this risk blind zone target is used as the state information of the environmental target.

[0048] As can be seen, this application can obtain the current predicted state information of the target in the risk blind zone by inversely solving the state information of the predicted collision point of the target in the risk blind zone, and thus use it as the state information of the environmental target. On the one hand, the state information of each environmental target in the environmental model can be visualized and displayed. On the other hand, as mentioned above, the starting point of the predicted trajectory set can be calibrated to a real and reasonable position that meets the collision constraints.

[0049] It should be noted that the above Figure 1 The method shown in the embodiment is executed cyclically at a preset frequency to achieve continuous tracking and real-time risk response of dynamic targets within the blind zone. The preset frequency can be 10-100Hz. That is, the following steps are executed cyclically at the preset frequency: Obtain a first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot area; for each blind spot target, based on the first predicted trajectory set corresponding to the blind spot target and the vehicle's target planning trajectory, determine whether there is a collision risk between the blind spot target and the vehicle; based on the determination result, determine the target trajectory set corresponding to the risky blind spot targets with collision risk from the first predicted trajectory set; generate collision information based on the target trajectory set and multiple candidate planning trajectories of the vehicle; generate a first control command based on the collision information, and control the vehicle's driving state based on the first control command; stop the loop until a preset loop stop condition is met.

[0050] The preset loop stopping condition includes at least one of the following: The vehicle has passed through the blind spot risk section of the road; The onboard unit determines that the vehicle's planned trajectory and the first predicted trajectory sets corresponding to all blind spot targets do not pose a collision risk. In one embodiment, obtaining the first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot includes: sending an information acquisition request to a roadside unit so that the roadside unit can collect the current state information of each blind spot target among the multiple blind spot targets; receiving the current state information of each blind spot target fed back by the roadside unit; and, for each blind spot target, performing trajectory prediction on the blind spot target based on the current state information of the blind spot target to obtain the corresponding first predicted trajectory set.

[0051] In this embodiment, the roadside unit is generally deployed at blind spot intersections and is equipped with sensors such as cameras and lidar to detect targets in the vehicle's blind spot, such as pedestrians and other vehicles.

[0052] When a vehicle is in motion, it uses high-precision maps and its own location information to determine that it is about to enter or is currently in a blind spot risk section. In this case, it sends an information request to the roadside unit in the blind spot risk section. Blind spot risk sections include: intersections, sharp bends, hill crests, road construction areas, and sections of road adjacent to large vehicles.

[0053] After receiving an information acquisition request, the roadside unit responds to the request by using the current status information of the dynamic target in the blind spot sensed by the sensors on the roadside unit, and sends the current status information of the dynamic target to the vehicle unit. Current status information includes: real-time location, speed, and type. For an explanation of the type, please refer to the corresponding description in the above embodiments, which will not be repeated here. Based on the current state information of the dynamic target and the additional environmental information that the roadside unit may provide, the vehicle-mounted unit runs a trajectory prediction algorithm to collaboratively predict the future movement of the dynamic target in the blind spot, and obtains the first predicted trajectory set of each dynamic target; wherein, the first predicted trajectory set of each dynamic target has a corresponding time to collision (TTC). Additional environmental information includes: the location of traffic lights in blind spots, and a topological map of the intersection.

[0054] In this application, the starting point state information obtained by the above reverse solution (i.e., the current predicted state information of the target in the risk blind zone) is different from or incompatible with the configuration parameters of the real target state information collected by the sensors on the vehicle. Therefore, the vehicle cannot directly process the predicted state information of the target in the risk blind zone. It is necessary to first convert the signal carrying the predicted state information into a virtual sensor signal. Specifically: In one embodiment, generating the state information of the environmental target based on the predicted state information includes: obtaining target configuration parameters; the target configuration parameters are configuration parameters of the state information of the real target collected by the sensors in the vehicle; setting the configuration parameters of the predicted state information as the target configuration parameters to obtain the state information of the environmental target; the target configuration parameters include: target format, transmission interface and acquisition frequency.

[0055] Converting a signal carrying predicted state information of a target in the blind zone into a virtual sensor signal includes: configuring at least one of the target format, transmission interface, and acquisition frequency of the predicted state information in the signal to be consistent with at least one of the format, transmission interface, and acquisition frequency of the state information of the real target in the real signal; and encapsulating the configured predicted state information into a virtual sensor signal that conforms to the vehicle's internal perception interface specification.

[0056] The vehicle's existing real sensors include millimeter-wave radar and / or lidar; virtual sensor signals can be virtual lidar point cloud signals or virtual millimeter-wave radar signals.

[0057] The format, transmission interface, and acquisition frequency of the virtual sensor signal are fully compatible with or highly similar to the configuration parameters of the output signals of the vehicle's existing real sensors.

[0058] It is evident that the predicted state information of the blind spot target is generated based on the state information of the estimated collision point on a certain predicted trajectory in the first predicted trajectory set. The state information of each trajectory point in the first predicted trajectory set is generated based on the state information of the blind spot target collected by the roadside unit. Therefore, the predicted state information of the blind spot target is actually consistent with the configuration parameters of the blind spot target's state information. This application converts the predicted state information of the blind spot target into a virtual sensor signal. The configuration information of the virtual sensor signal is consistent with the real signal collected by the vehicle. This allows the predicted state information of the blind spot target to be injected into the vehicle unit in the form of native perception data of the vehicle unit. There is no need to deeply reconstruct the mature planning and control algorithm, which greatly reduces the system integration complexity, development cost and verification cycle, and is conducive to the rapid promotion of the technology.

[0059] In one embodiment, the above-mentioned environmental model includes not only environmental targets corresponding to blind zone targets, but also environmental targets corresponding to real targets, specifically: The vehicle collects state information of multiple real targets within its detectable range through sensors installed on the vehicle. The detectable range of the vehicle refers to the environmental range that the vehicle can detect through its sensors. For each real target, an environmental model is constructed based on the state information of the real target, the state information of each environmental target corresponding to each risk blind zone target, and the first identifier.

[0060] Based on the state information of the real target, an environment model is constructed, specifically including: generating the state information of the corresponding environmental target based on the state information of the real target; and constructing an environment model based on the state information of the corresponding environmental target. The environmental target includes the environmental target corresponding to each real target, as well as the state information of each environmental target. Among them, generating environmental target state information based on the state information of the real target includes: using the state information of the real target as the state information of the environmental target.

[0061] Among them, an environmental model is constructed based on the state information of environmental targets corresponding to real targets and the state information of environmental targets corresponding to risk blind zone targets, including: The state information of environmental targets corresponding to real targets and the state information of environmental targets corresponding to risk blind zone targets are spatiotemporally synchronized and processed with a coordinate system. Based on the processed state information of environmental targets corresponding to real targets and environmental targets corresponding to risk blind zone targets, an environmental model is constructed.

[0062] The coordinate system processing includes converting the position coordinates in the state information of either the state information of the environmental target corresponding to the real target or the state information of the environmental target corresponding to the risk blind zone target into position coordinates in a unified coordinate system.

[0063] As can be seen, this application can construct an environment model with a coordinate system that includes the state information of real targets within the vehicle's detectable range and the current predicted state information of targets in the blind spot.

[0064] In one embodiment, the collision information is the duration between the current time and the time when the vehicle collides. The step of generating collision information based on the second predicted trajectory set and the candidate planned trajectory includes: for each environmental target, determining whether there is a collision risk between the environmental target and the vehicle based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle; if there is a collision risk, determining the duration between the current time and the time when the vehicle collides with the environmental target; and generating the collision information based on the duration corresponding to each environmental target.

[0065] For each environmental target, if there is at least one predicted trajectory in the second predicted trajectory set corresponding to the environmental target and at least one candidate planned trajectory among the multiple candidate planned trajectories of the vehicle (i.e., a collision point), then it is determined that there is a collision risk between the environmental target and the vehicle; if there is a collision risk, then the time between the current moment and the moment when the vehicle collides with the environmental target is determined. In one embodiment, for each environmental target, determining the duration between the current time and the time of collision includes: determining the target prediction trajectory in the second predicted trajectory set of the environmental target that collided, and the pre-collision planning trajectory among multiple candidate planning trajectories; determining the actual collision point between the target prediction trajectory and the pre-collision planning trajectory; and determining the duration between the current time and the time of collision based on the actual collision point, the target prediction trajectory, and the pre-collision planning trajectory.

[0066] For each collision point, determine the target predicted trajectory in the second predicted trajectory set where the collision occurred, as well as the pre-collision planned trajectory among multiple candidate planned trajectories. Determine the intersection of the target predicted trajectory and the pre-collision planned trajectory, which is the actual collision point. Determine the time required for the vehicle to travel to the actual collision point according to the candidate planned trajectory, which is the time between the current moment and the moment of collision. The minimum duration is determined from the durations corresponding to each actual collision point, and is taken as the duration corresponding to the target in that environment.

[0067] Collision information is generated based on the duration corresponding to each environmental target. This collision information specifically includes: the duration from the current moment to the moment the vehicle collides, and the corresponding minimum allowable distance (MPD). The minimum allowable distance is the distance from the vehicle's current position to the point of collision.

[0068] In one embodiment, generating the collision information based on the duration corresponding to each environmental target includes: determining the minimum duration among the durations corresponding to each environmental target to obtain the collision information.

[0069] The minimum duration among the durations corresponding to each environmental target is used as the collision information; Alternatively, determine the MPD (Minimum Time Limit) between the vehicle's current location and the collision point corresponding to the minimum time, and use the MPD and the minimum time to form the collision information.

[0070] In one embodiment, generating the first control command based on the collision information includes: If the minimum duration in the collision information is not greater than the first preset duration, a first driving control command is generated. The first driving control command includes a braking command or an emergency steering command. The emergency steering command is used to indicate that the steering angle is greater than the preset steering angle within a preset unit time. If the minimum duration in the collision information is greater than the first preset duration and less than the second preset duration, then based on the target predicted trajectory and the pre-collision planned trajectory, a final planned trajectory is determined, and a second driving control command is generated based on the final planned trajectory; the second driving control command includes a deceleration command and / or a smooth steering command, the smooth steering command being used to indicate that the steering angle is less than a preset steering angle within a preset unit time.

[0071] In this embodiment, if the minimum duration in the collision information is not greater than the first preset duration, it indicates an emergency situation, a braking command is generated, and the vehicle is controlled to brake urgently based on the braking command. A first driving control command is generated, which includes a braking command or an emergency steering command. The steering intensity of the emergency steering command is greater than that of the smooth steering command. Specifically, the emergency steering command is used to indicate that the steering angle is greater than the preset steering angle within a preset unit time, while the smooth steering command is used to indicate that the steering angle is less than the preset steering angle within a preset unit time. Therefore, compared with the smooth steering command, the emergency steering command has a larger steering angle within the preset unit time, and thus a greater steering intensity.

[0072] If the minimum duration in the collision information is greater than the first preset duration and less than the second preset duration, it indicates that there is still a certain amount of time before the collision. Based on the target predicted trajectory and the pre-collision planned trajectory, the final planned trajectory is determined, specifically including: Based on the predicted trajectory of the target that will collide and the pre-collision planned trajectory, a new trajectory that will not collide is replanned. Alternatively, based on the pre-collision planning trajectories among multiple candidate planning trajectories that will collide, determine the planning trajectories that will not collide among the multiple candidate planning trajectories, and select one planning trajectory from the planning trajectories that will not collide as the final planning trajectory.

[0073] A second driving control command is generated based on the final planned trajectory, so that the vehicle can be controlled to drive according to the final planned trajectory based on the second driving control command.

[0074] In this embodiment, when a collision risk exists, the on-board unit automatically generates driving control commands and sends them to the vehicle's execution unit for execution, enabling automatic risk avoidance without driver intervention. The vehicle's execution unit includes: electronic stability program (ESP) and / or electric power steering (EPS).

[0075] In one embodiment, after determining whether there is a collision risk between the environmental target and the vehicle based on the second predicted trajectory set corresponding to each environmental target and the candidate planned trajectory of the vehicle, the method further includes: if there is no collision risk between all environmental targets and the vehicle, then visually displaying the state information of each environmental target in the environmental model; and generating a second control command based on the state information of each environmental target, and controlling the driving state of the vehicle based on the second control command.

[0076] If there is no collision risk between any environmental target and the vehicle, the aforementioned driving control commands will not be automatically generated to achieve automatic risk avoidance. Therefore, this application can also visualize the state information of each environmental target in the environmental model, allowing users to see the real-time state of each environmental target. Simultaneously, based on the state information and / or type of each environmental target, a second control command is generated and sent to the vehicle's execution unit for execution.

[0077] In this process, a second control command is generated based on the real-time state of each environmental target in the environmental model, and the vehicle's driving state is controlled based on the second control command. Although the confidence level of this decision is lower than that of the decision based on the first control command to control the vehicle's driving state, controlling the vehicle's driving state based on the second control command can still achieve automatic risk avoidance to a certain extent.

[0078] The second control command generated varies depending on the type of blind target. For example, the second control command generated is different for vehicles and pedestrians.

[0079] Reference Figure 2 The vehicle-mounted unit of this application includes: V2X communication module: used for information exchange with roadside units.

[0080] The on-board processing unit specifically includes an environmental perception module, a planning and control module, and a data processing and prediction module. Data processing and prediction module: used to receive the current status information of blind targets sent by the roadside unit, predict the trajectory of blind targets, and generate virtual sensing signals.

[0081] Environmental perception module: used to synchronize virtual perception signals with real sensor signals in time and space and unify coordinate systems to build a unified environmental model.

[0082] Planning and Control Module: Used to perform risk assessment based on the environmental model and the predicted trajectory associated with environmental targets in the environmental model, and to generate control commands to control the driving status of the vehicle.

[0083] Actuator: Used to execute control commands issued by the planning and control module.

[0084] The position and speed of blind targets may change. The roadside unit monitors and updates the current status information of the blind target in real time and feeds back the vehicle status information to the vehicle unit. The vehicle unit then performs the steps of dynamically adjusting the vehicle's driving status control at a frequency of 10 to 100 Hz until the risk removal condition (i.e. the above-mentioned loop stop condition) is met, the normal driving model is restored, and the blind spot information request is stopped from being sent to the vehicle unit. The steps for controlling the vehicle's driving status include: When the vehicle-mounted unit determines that the vehicle is approaching a blind spot section based on high-precision map positioning, it sends a blind spot information request (i.e., the aforementioned information acquisition request) to the roadside unit. The roadside unit perceives blind spot targets through cameras, LiDAR, or millimeter-wave radar, and obtains the target state information of each blind spot target (i.e., the aforementioned current state information of the blind spot target). The target state information includes: position, speed, type, and direction. Based on the current state information of the blind spot targets and the roadside information, the vehicle-mounted processing unit predicts the first predicted trajectory set of the blind spot targets through multi-model fusion prediction, and calculates the TTC corresponding to each first predicted trajectory set. The vehicle-mounted processing unit determines the target trajectory set with collision risk in the first predicted trajectory set, determines the estimated collision point between the target trajectory set and multiple candidate planned trajectories of the vehicle, and reversely solves the current predicted state information of the blind spot target based on the estimated collision point. Based on the predicted state information, a virtual signal (corresponding to the aforementioned virtual sensor signal) is generated. The virtual signal can specifically be a virtual LiDAR point cloud signal. The target format, transmission interface, and acquisition frequency of the virtual signal are configured to be consistent with the format, transmission interface, and acquisition frequency of the real signal. (Corresponding to the virtual camera image in the figure, compatible with real sensor format), construct an environment model, which integrates virtual and real signals. The environment model is a complete environment perception model. The planning and control module performs risk assessment and decision-making, collision risk analysis (i.e., based on the target trajectory set and multiple candidate planning trajectories of the vehicle, generate collision information, generate a first control command based on the collision information, and control the driving state of the vehicle based on the first control command), avoidance strategy selection, and trajectory planning (i.e., if the minimum duration in the collision information is not greater than a first preset duration, generate a first driving control command; if the minimum duration in the collision information is greater than the first preset duration and less than a second preset duration, determine the final planning trajectory based on the target predicted trajectory and the pre-collision planning trajectory, and generate a second driving control command based on the final planning trajectory). The generated control commands are sent to the execution unit, including automatic commands, steering commands, acceleration commands, or deceleration commands. The execution mechanism executes the corresponding commands to control the vehicle, so as to perform avoidance actions, maintain a safe distance, and stabilize the vehicle state.

[0085] Reference Figure 3 The overall scheme of this application is briefly described as follows: S1. Obtain the current status information of the target in the blind spot; S2, Collaborative Prediction; Predict the trajectory of the target in the blind spot and output the predicted trajectory set, namely the first predicted trajectory set mentioned above.

[0086] S3, Virtual signal generation; Based on the first predicted trajectory set of blind spot targets with collision risk and the predicted collision point of the vehicle's target planned trajectory, the trajectory is solved in reverse based on the estimated collision point to predict the state information of the starting point of the predicted trajectory set, i.e. the current predicted state information of the blind spot target; state points are generated, including position, speed, and type; the state points are encapsulated to obtain a virtual perception signal, which carries the identifier of the blind spot target. S4, Environment Model Construction; Create or update virtual target entries in the environment model. Virtual target entries include: state points, unique identification (ID) (which is obtained from the identification of blind zone targets), and pointers to associated predicted trajectories.

[0087] S5. Planning and Decision-Making.

[0088] Scenario 1: Using the unique identifier under the virtual target entry, find the first predicted trajectory set corresponding to the unique identifier from the first predicted trajectory set of multiple blind spot targets. Compare each found first predicted trajectory set with the vehicle's trajectory in time and space to determine whether there is a collision risk. If there is a collision risk, calculate the precise TTC and minimum permissible distance (MPD) and make a decision based on TTC and MPD.

[0089] Scenario 2: If there is no collision risk between all the first predicted trajectory sets and the vehicle's trajectory, the instantaneous position and type of the target in the environment model are visualized, and a decision is made based on the instantaneous position and type of the target in the environment model. The confidence level of this decision is lower than that of the decision in Scenario 1.

[0090] As can be seen, this application improves the obstacle avoidance of blind spot targets with collision risk from being taken over by the driver to actively controlling the vehicle through control commands, that is, upgrading from "passive warning" to "active closed-loop control", which can automatically execute avoidance actions and avoid situations where the driver is not able to react in time.

[0091] Based on the same concept, embodiments of the present invention also provide a vehicle driving control device. Figure 4 This is a structural block diagram of a vehicle driving control device provided in an embodiment of this application, such as... Figure 4 As shown, the device 400 includes: The acquisition module 401 is used to acquire a first set of predicted trajectories corresponding to multiple blind spot targets within the vehicle's blind spot; the first set of predicted trajectories includes at least one predicted trajectory. The judgment module 402 is used to determine whether there is a collision risk between the blind spot target and the vehicle based on the first predicted trajectory set corresponding to the blind spot target and the target planned trajectory of the vehicle for each blind spot target; the target planned trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planned trajectories, and the multiple candidate planned trajectories are determined based on the state information of the vehicle; The determination module 403 is used to determine, based on the judgment result, the target trajectory set corresponding to the risk blind zone targets with collision risk from the first predicted trajectory set; The generation module 404 is used to generate collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle; The control module 405 is used to generate a first control command based on the collision information, and to control the driving state of the vehicle based on the first control command.

[0092] In one embodiment, the generation module 404 is specifically used for: Construct an environment model; the environment model includes attribute information of various environmental targets in the current environment in which the vehicle is located, which pose a collision risk to the vehicle, and the attribute information includes a first identifier and state information; Associate each target trajectory set with the corresponding environmental target in the environmental model; The target trajectory set associated with each environmental target is respectively used as the second predicted trajectory set of the environmental target; Collision information is generated based on the second predicted trajectory set and the candidate planned trajectory.

[0093] In one embodiment, the generation module 404, when constructing the environment model, is specifically used for: For each risk blind zone target, determine the estimated collision points between the target trajectory set corresponding to the risk blind zone target and the target planned trajectory; Based on the state information of the estimated collision point, the state information of the environmental target is generated; and based on the second identifier of the risk blind zone target, the first identifier of the environmental target is generated. The environmental model is constructed based on the state information of each environmental target and the first identifier.

[0094] In one embodiment, when the generation module 404 associates each set of target trajectories with the corresponding environmental targets in the environment model, it is specifically used for: For each environmental target in the environmental model, a matching target identifier is determined from a plurality of second identifiers based on the first identifier of the environmental target; Determine the first predicted trajectory set of the blind zone target indicated by the target identifier; Associate the first predicted trajectory set of the blind zone target indicated by the target identifier with the environmental target.

[0095] In one embodiment, when generating state information of the environmental target based on the estimated collision point state information, the generation module 404 is specifically used for: For each target in the risk blind zone, the current predicted state information of the target in the risk blind zone is determined based on the state information of the predicted collision point; Based on the predicted state information, the state information of the environmental target is generated.

[0096] In one embodiment, the collision information is the duration between the current time and the time corresponding to the vehicle collision. When generating the collision information based on the second predicted trajectory set and the candidate planned trajectory, the generation module 404 is specifically used for: For each environmental target, based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle, it is determined whether there is a collision risk between the environmental target and the vehicle; If there is a risk of collision, determine the current moment and the duration between the moment the vehicle collides with the environmental target; The collision information is generated based on the duration corresponding to each environmental target.

[0097] In one embodiment, when determining the duration between the current moment and the moment of collision, the generation module 404 is specifically used for: Determine the target prediction trajectory in the second predicted trajectory set of the environmental target that collided, as well as the pre-collision planned trajectory among multiple candidate planned trajectories; Indeed, the actual collision point between the predicted target trajectory and the pre-collision planned trajectory; Based on the actual collision point, the target predicted trajectory, and the pre-collision planned trajectory, the duration between the current moment and the moment of collision is determined.

[0098] In one embodiment, the generation module 404 generates the collision information based on the duration corresponding to each environmental target, including: The minimum duration among the durations corresponding to each environmental target is determined to obtain collision information.

[0099] In one embodiment, when the control module 405 generates a first control command based on the collision information, it is specifically used for: If the minimum duration in the collision information is not greater than the first preset duration, a first driving control command is generated. The first driving control command includes a braking command or an emergency steering command. The emergency steering command is used to indicate that the steering angle is greater than the preset steering angle within a preset unit time. If the minimum duration in the collision information is greater than the first preset duration and less than the second preset duration, then based on the target predicted trajectory and the pre-collision planned trajectory, a final planned trajectory is determined, and a second driving control command is generated based on the final planned trajectory; the second driving control command includes a deceleration command and / or a smooth steering command, the smooth steering command being used to indicate that the steering angle is less than a preset steering angle within a preset unit time.

[0100] In one embodiment, the device further includes a visualization module, used after the generation module 404 determines whether there is a collision risk between the environmental target and the vehicle based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle; If there is no risk of collision between any environmental target and the vehicle, the state information of each environmental target in the environmental model is visualized; and a second control command is generated based on the state information of each environmental target, and the driving state of the vehicle is controlled based on the second control command.

[0101] In one embodiment, when generating the state information of the environmental target based on the predicted state information, the generation module 404 is specifically used for: Obtain target configuration parameters; the target configuration parameters are the configuration parameters of the state information of the real target collected by the sensors in the vehicle. The configuration parameters of the predicted state information are set to the target configuration parameters to obtain the state information of the environmental target; The target configuration parameters include: target format, transmission interface, and acquisition frequency.

[0102] In one embodiment, the acquisition module 401 is specifically used for: Send an information acquisition request to the roadside unit so that the roadside unit can collect the current status information of each blind spot target among the multiple blind spot targets; Receive the current status information of each blind spot target fed back by the roadside unit; For each blind zone target, trajectory prediction is performed on the blind zone target based on the current state information of the blind zone target to obtain the corresponding first predicted trajectory set.

[0103] It is understood that the device provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0104] Reference Figure 5 The present invention also provides an electronic device, which may include a processor 501 and a memory 501, wherein the processor 501 and the memory 501 can communicate with each other through a bus or other means.

[0105] The processor 502 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application, or it may be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or other chips, or combinations of the above types of chips.

[0106] Memory 501 may include mass storage for data or instructions. For example, and not limitingly, memory may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory may include removable or non-removable (or fixed) media. Where appropriate, memory may be internal or external to an electronic device. In a particular embodiment, memory may be non-volatile solid-state memory.

[0107] In one instance, memory 501 may be read-only memory (ROM). In one instance, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0108] The processor 502 implements any of the vehicle driving control methods described in the above embodiments by reading and executing computer program instructions stored in the memory.

[0109] In one example, the electronic device may further include a communication interface and a bus. The processor, memory, and communication interface are connected via the bus to communicate with each other. The communication interface is primarily used to enable communication between the various modules, devices, units, and / or equipment in the embodiments of this application. Where appropriate, the bus may include one or more buses.

[0110] Furthermore, in conjunction with the vehicle driving control methods in the above embodiments, this invention can be implemented using a computer-readable storage medium. This computer-readable storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the vehicle driving control methods in the above embodiments.

[0111] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The storage medium can be read-only memory (ROM), random access memory (RAM), flash memory, hard disk drive (HDD), or solid-state drive (SSD), etc.; the storage medium can also include combinations of the above types of memory.

[0112] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0113] Similarly, it should be understood that, in order to simplify this disclosure and aid in understanding one or more of the various aspects of the invention, in the above description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof. However, this method of disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, inventive aspects lie in fewer than all features of a single foregoing disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into this detailed description, wherein each claim itself is a separate embodiment of the invention.

[0114] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claims. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

Claims

1. A vehicle driving control method, characterized in that, The method includes: Obtain a first set of predicted trajectories corresponding to multiple blind spot targets within the vehicle's blind spot; the first set of predicted trajectories includes at least one predicted trajectory. For each blind spot target, based on the first predicted trajectory set corresponding to the blind spot target and the target planned trajectory of the vehicle, it is determined whether there is a collision risk between the blind spot target and the vehicle; the target planned trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planned trajectories, and the multiple candidate planned trajectories are determined based on the state information of the vehicle; Based on the judgment results, the target trajectory sets corresponding to the risk blind zone targets with collision risk are determined from the first predicted trajectory set; Based on the target trajectory set and multiple candidate planned trajectories of the vehicle, collision information is generated; A first control command is generated based on the collision information, and the driving state of the vehicle is controlled based on the first control command.

2. The method according to claim 1, characterized in that, The generation of collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle includes: Construct an environment model; the environment model includes attribute information of various environmental targets in the current environment in which the vehicle is located, which pose a collision risk to the vehicle, and the attribute information includes a first identifier and state information; Associate each target trajectory set with the corresponding environmental target in the environmental model; The target trajectory set associated with each environmental target is respectively used as the second predicted trajectory set of the environmental target; Collision information is generated based on the second predicted trajectory set and the candidate planned trajectory.

3. The method according to claim 2, characterized in that, The construction environment model includes: For each risk blind zone target, determine the estimated collision points between the target trajectory set corresponding to the risk blind zone target and the target planned trajectory; Based on the state information of the estimated collision point, the state information of the environmental target is generated; and based on the second identifier of the risk blind zone target, the first identifier of the environmental target is generated. The environmental model is constructed based on the state information of each environmental target and the first identifier.

4. The method according to claim 3, characterized in that, The step of associating each target trajectory set with the corresponding environmental target in the environmental model includes: For each environmental target in the environmental model, a matching target identifier is determined from a plurality of second identifiers based on the first identifier of the environmental target; Determine the first predicted trajectory set of the blind zone target indicated by the target identifier; Associate the first predicted trajectory set of the blind zone target indicated by the target identifier with the environmental target.

5. The method according to claim 3, characterized in that, The step of generating environmental target state information based on the estimated collision point state information includes: For each target in the risk blind zone, the current predicted state information of the target in the risk blind zone is determined based on the state information of the predicted collision point; Based on the predicted state information, the state information of the environmental target is generated.

6. The method according to any one of claims 2 to 5, characterized in that, The collision information is the duration between the current time and the time corresponding to the vehicle collision. Generating collision information based on the second predicted trajectory set and the candidate planned trajectory includes: For each environmental target, based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle, it is determined whether there is a collision risk between the environmental target and the vehicle; If there is a risk of collision, determine the current moment and the duration between the moment the vehicle collides with the environmental target; The collision information is generated based on the duration corresponding to each environmental target.

7. The method according to claim 6, characterized in that, Determining the duration between the current moment and the moment of collision includes: Determine the target prediction trajectory in the second predicted trajectory set of the environmental target that collided, as well as the pre-collision planned trajectory among multiple candidate planned trajectories; Indeed, the actual collision point between the predicted target trajectory and the pre-collision planned trajectory; Based on the actual collision point, the target predicted trajectory, and the pre-collision planned trajectory, the duration between the current moment and the moment of collision is determined.

8. The method according to claim 6, characterized in that, The generation of collision information based on the duration corresponding to each environmental target includes: The minimum duration among the durations corresponding to each environmental target is determined to obtain collision information.

9. The method according to claim 5, characterized in that, The generation of the first control command based on the collision information includes: If the minimum duration in the collision information is not greater than the first preset duration, a first driving control command is generated. The first driving control command includes a braking command or an emergency steering command. The emergency steering command is used to indicate that the steering angle is greater than the preset steering angle within a preset unit time. If the minimum duration in the collision information is greater than the first preset duration and less than the second preset duration, then based on the target predicted trajectory and the pre-collision planned trajectory, a final planned trajectory is determined, and a second driving control command is generated based on the final planned trajectory; the second driving control command includes a deceleration command and / or a smooth steering command, the smooth steering command being used to indicate that the steering angle is less than a preset steering angle within a preset unit time.

10. The method according to claim 6, characterized in that, After determining whether there is a collision risk between the environmental target and the vehicle based on the second predicted trajectory set corresponding to the environmental target and the candidate planned trajectory of the vehicle, the method further includes: If there is no risk of collision between any environmental target and the vehicle, the state information of each environmental target in the environmental model is visualized; and a second control command is generated based on the state information of each environmental target, and the driving state of the vehicle is controlled based on the second control command.

11. The method according to claim 5, characterized in that, The step of generating state information of the environmental target based on the predicted state information includes: Obtain target configuration parameters; the target configuration parameters are the configuration parameters of the state information of the real target collected by the sensors in the vehicle. The configuration parameters of the predicted state information are set to the target configuration parameters to obtain the state information of the environmental target; The target configuration parameters include: target format, transmission interface, and acquisition frequency.

12. The method according to any one of claims 2 to 5, characterized in that, The step of obtaining the first predicted trajectory set corresponding to multiple blind spot targets within the vehicle's blind spot includes: Send an information acquisition request to the roadside unit so that the roadside unit can collect the current status information of each blind spot target among the multiple blind spot targets; Receive the current status information of each blind spot target fed back by the roadside unit; For each blind zone target, trajectory prediction is performed on the blind zone target based on the current state information of the blind zone target to obtain the corresponding first predicted trajectory set.

13. A vehicle driving control device, characterized in that, The device includes: The acquisition module is used to acquire a first set of predicted trajectories corresponding to multiple blind spot targets within the vehicle's blind spot; the first set of predicted trajectories includes at least one predicted trajectory. The judgment module is used to determine whether there is a collision risk between the blind spot target and the vehicle based on the first predicted trajectory set corresponding to the blind spot target and the target planned trajectory of the vehicle for each blind spot target; the target planned trajectory is the trajectory along the current driving direction of the vehicle among multiple candidate planned trajectories, and the multiple candidate planned trajectories are determined based on the state information of the vehicle; The determination module is used to determine, based on the judgment result, the target trajectory set corresponding to the risk blind zone targets with collision risk from the first predicted trajectory set; The generation module is used to generate collision information based on the target trajectory set and multiple candidate planned trajectories of the vehicle; The control module is used to generate a first control command based on the collision information, and to control the driving state of the vehicle based on the first control command.

14. An electronic device, characterized in that, include: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, the processor executing the computer instructions to perform the method of any one of claims 1-12.

15. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the method of any one of claims 1-12.

Images