Vehicle driving strategy determination method, vehicle mixed driving control method and unmanned vehicle

Abstract

The invention provides a vehicle driving strategy determination method, a vehicle mixed driving control method and an unmanned vehicle, and relates to the field of unmanned driving, automatic driving and unmanned vehicles. The vehicle driving strategy determination method comprises the steps that under the condition that a collision risk exists between a target vehicle and surrounding objects, at least one collision scheme and at least one risk avoiding scheme are generated, the collision scheme refers to a collision coping mode adopted when the target vehicle cannot avoid collision, and the risk avoiding scheme refers to a collision coping mode adopted when the target vehicle cannot avoid collision. The risk avoiding scheme refers to an avoiding coping mode adopted by the target vehicle to avoid collision; calculating the collision cost of each collision scheme, and calculating the risk cost of each risk avoiding scheme; and determining a target driving strategy from the at least one collision scheme and the at least one risk avoiding scheme based on the collision cost of each collision scheme and the risk cost of each risk avoiding scheme.

Description

Technical Field

[0001] This application relates to the technical fields of driverless, autonomous driving, and driverless vehicle technologies, and particularly relates to a method for determining a vehicle driving strategy, a method for controlling vehicle mixed traffic, and a driverless vehicle. Background Art

[0002] A vehicle may pose safety risks to the surrounding environment and itself due to its own reasons or software and hardware failures, and may even directly cause safety accidents.

[0003] In traditional solutions, when a vehicle may collide with the surrounding environment, the vehicle actions are usually optimized based on the potential losses generated after the collision to reduce the collision damage. Summary of the Invention

[0004] In view of this, embodiments of this application provide a method for determining a vehicle driving strategy, a method for controlling vehicle mixed traffic, and a driverless vehicle.

[0005] In a first aspect, an embodiment of this application provides a method for determining a vehicle driving strategy, including: when there is a collision risk between a target vehicle and surrounding objects, generating at least one collision plan and at least one avoidance plan, where a collision plan refers to a collision response method taken when the target vehicle cannot avoid a collision, and an avoidance plan refers to an avoidance response method taken by the target vehicle to avoid a collision; calculating the collision cost of each collision plan and calculating the risk cost of each avoidance plan; and determining a target driving strategy from at least one collision plan and at least one avoidance plan based on the collision cost of each collision plan and the risk cost of each avoidance plan.

[0006] In combination with the first aspect, in some implementation manners of the first aspect, when there is a collision risk between a target vehicle and surrounding objects, generating at least one collision plan and at least one avoidance plan includes: when it is determined that there is a collision risk between the target vehicle and surrounding objects, generating at least one avoidance plan; and generating corresponding collision plans based on at least one avoidance plan.

[0007] In combination with the first aspect, in some implementation manners of the first aspect, generating at least one avoidance plan includes: obtaining the motion information of the target vehicle and the state data of surrounding objects; generating a candidate avoidance sequence based on the motion information of the target vehicle and the state data of surrounding objects, where the candidate avoidance sequence includes a combination of one or more actions among steering, braking, and acceleration; and performing a feasibility check on each candidate avoidance sequence, and determining the candidate avoidance sequence that passes the feasibility check as an avoidance plan.

[0008] In combination with the first aspect, in some implementations of the first aspect, based on at least one risk avoidance plan, a corresponding collision plan is generated, including: for each risk avoidance plan, predicting the execution process of the risk avoidance plan; if the execution of the risk avoidance plan fails, determining the potential collision object and / or potential collision position after the execution of the risk avoidance plan, where the failure of execution means that the target vehicle cannot avoid a collision with surrounding objects; based on the potential collision object and / or potential collision position after the execution of the risk avoidance plan, generating the collision plan corresponding to the risk avoidance plan, and the associated information of the collision plan includes the collision object and / or the collision position.

[0009] In combination with the first aspect, in some implementations of the first aspect, the collision plan corresponding to the risk avoidance plan includes a collision attitude optimization instruction and / or a collision energy distribution strategy; based on the potential collision object and / or potential collision position after the execution of the risk avoidance plan, generating the collision plan corresponding to the risk avoidance plan, including: based on the type of the potential collision object and / or the potential collision position, generating a collision attitude optimization instruction, where the collision attitude optimization instruction is used to indicate collision position optimization and / or collision angle optimization; and / or, based on the distribution of the potential collision position and the buffer of the target vehicle, generating a collision energy distribution strategy, and the collision energy distribution strategy includes a distribution rule for dispersing the impact energy generated by the collision to the buffer according to a preset ratio.

[0010] In combination with the first aspect, in some implementations of the first aspect, before generating at least one collision plan and at least one risk avoidance plan when there is a collision risk between the target vehicle and surrounding objects, it further includes: Obtaining the motion information of the target vehicle and the status data of surrounding objects; based on the motion information of the target vehicle, generating at least one predicted trajectory of the target vehicle at a future target time; if based on the status data of surrounding objects, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects, then it is determined that there is a collision risk between the target vehicle and the surrounding object; where the motion information includes at least one of pose, speed, acceleration, kinematic parameters, and dynamic parameters; and the status data includes at least one of position, speed, size, and motion direction.

[0011] In combination with the first aspect, in some implementations of the first aspect, if based on the status data of surrounding objects, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects, it includes: based on the status data of surrounding objects, calculating the spatio-temporal overlap probability between the predicted trajectory of the target vehicle and the surrounding objects; if the spatio-temporal overlap probability is greater than a preset probability threshold, then it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects.

[0012] In combination with the first aspect, in some implementations of the first aspect, calculating the collision cost of each collision scenario includes: for each collision scenario, through at least one cost calculation function in the collision cost model, based on the costs corresponding to the collision objects and / or collision positions associated with the collision scenario, and the cost corresponding to the target vehicle, calculating the collision cost of the collision scenario; wherein, the cost of the collision object is determined based on the type of the collision object or the value of the collision object.

[0013] In combination with the first aspect, in some implementations of the first aspect, calculating the risk cost of each avoidance scenario includes: for each avoidance scenario, through at least one cost calculation function in the risk cost model, based on at least one of the type of the avoidance action, the state data of the surrounding objects, the motion information of the target vehicle, and the environmental parameters where the target vehicle is located, calculating the risk cost of the avoidance scenario, wherein the risk cost is the expected cost of the occurrence probability of a risk event during the execution of the avoidance scenario.

[0014] In combination with the first aspect, in some implementations of the first aspect, the motion information of the target vehicle includes dynamic parameters, the dynamic parameters include load and / or center of gravity height, and the environmental parameters where the target vehicle is located include road surface adhesion coefficient and / or visibility.

[0015] In combination with the first aspect, in some implementations of the first aspect, based on the collision costs of each collision scenario and the risk costs of each avoidance scenario, determining a target driving strategy from at least one collision scenario and at least one avoidance scenario includes: comparing the numerical values of the collision costs of the collision scenarios and the risk costs of the avoidance scenarios, and selecting the scenario with the smallest numerical value as the target driving strategy; or, selecting a set of scenarios that meet the safety threshold from the collision scenarios and the avoidance scenarios, and selecting the scenario with the lowest cost from the set of scenarios as the target driving strategy.

[0016] In a second aspect, an embodiment of the present application provides a vehicle mixed traffic control method, including: in the case where there is a collision risk between an autonomous vehicle and a human-driven vehicle, determining a target driving strategy corresponding to the autonomous vehicle, wherein the target driving strategy is obtained based on the method described in the first aspect; controlling the autonomous vehicle to drive based on the target driving strategy.

[0017] In a third aspect, an embodiment of the present application provides a vehicle driving strategy determination device, including: a generation module, configured to generate at least one collision plan and at least one avoidance plan when there is a collision risk between a target vehicle and surrounding objects, where the collision plan refers to a collision response method adopted when the target vehicle cannot avoid a collision, and the avoidance plan refers to an avoidance response method adopted by the target vehicle to avoid a collision; a calculation module, configured to calculate the collision cost of each collision plan and calculate the risk cost of each avoidance plan; a first determination module, configured to determine a target driving strategy from at least one collision plan and at least one avoidance plan based on the collision cost of each collision plan and the risk cost of each avoidance plan.

[0018] In a fourth aspect, an embodiment of the present application provides a vehicle mixed traffic control device, including: a second determination module, configured to determine a target driving strategy corresponding to an autonomous vehicle when there is a collision risk between the autonomous vehicle and a manned vehicle, where the target driving strategy is obtained based on the method described in the first aspect; a control module, configured to control the autonomous vehicle to drive based on the target driving strategy.

[0019] In a fifth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program for executing the method described in the first aspect and / or the second aspect.

[0020] In a sixth aspect, an embodiment of the present application provides an autonomous vehicle, including: a processor; a memory for storing instructions executable by the processor; the processor is configured to execute the method described in the first aspect and / or the second aspect.

[0021] In a seventh aspect, an embodiment of the present application provides a computer program product, which includes instructions that, when executed on an autonomous vehicle, cause the autonomous vehicle to implement the method described in the first aspect and / or the second aspect.

[0022] In the present application, first, at least one collision plan and at least one avoidance plan are generated simultaneously, providing multiple possibilities for subsequent selection of a better driving strategy. Then, the collision cost of each collision plan and the risk cost of each avoidance plan are calculated to intuitively show the potential losses and risks of different plans in a quantitative manner. Finally, based on cost comparison and comprehensive consideration, a target driving strategy is determined from the collision plan and the avoidance plan, fully considering the balance between safety and driving efficiency, and being able to flexibly select the most appropriate driving method in a complex traffic environment.

[0023] Overall, the present application realizes the forward-looking decision-making of the vehicle driving strategy, which not only helps to reduce the losses caused by collisions, but also can effectively avoid unnecessary risks, improves the vehicle's response ability when facing collision risks, enhances driving safety, and at the same time significantly improves the rationality and scientificity of the vehicle driving strategy, effectively reducing the probability of safety accidents and various possible losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] By describing the embodiments of the present application in more detail in conjunction with the drawings, the above and other objects, features, and advantages of the present application will become more obvious. The drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation to the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0025] Figure 1 The following is a schematic flowchart of a method for determining a vehicle driving strategy provided by an embodiment of the present application.

[0026] Figure 2 The following is a schematic flowchart of a method for generating at least one collision plan and at least one risk avoidance plan provided by an embodiment of the present application.

[0027] Figure 3 The following is a schematic flowchart of a method for determining a vehicle driving strategy provided by another embodiment of the present application.

[0028] Figure 4 The following is a schematic flowchart of a method for determining a target driving strategy provided by an embodiment of the present application.

[0029] Figure 5 The following is a schematic flowchart of a vehicle mixed traffic control method provided by an embodiment of the present application.

[0030] Figure 6 The following is a schematic structural diagram of a vehicle driving strategy determination device provided by an embodiment of the present application.

[0031] Figure 7 The following is a schematic structural diagram of a vehicle mixed traffic control device provided by an embodiment of the present application.

[0032] Figure 8 The following is a schematic structural diagram of a driverless vehicle provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0034] Figure 1 The following is a schematic flowchart of a method for determining a vehicle driving strategy provided by an embodiment of the present application. Exemplarily, as Figure 1 shown, the method includes the following steps.

[0035] Step S110, when there is a collision risk between the target vehicle and surrounding objects, generate at least one collision plan and at least one avoidance plan.

[0036] Surrounding objects refer to other traffic participants within a certain range around the target vehicle. Exemplarily, surrounding objects include other vehicles, pedestrians, obstacles, etc.

[0037] Optionally, the target vehicle needs to use various sensors, such as cameras, lidar, etc., to continuously sense the information of the surrounding environment, including the position, speed, acceleration, type, etc. of the surrounding objects. According to the sensed information of the surrounding objects, evaluate the collision risk between the target vehicle and the surrounding objects. For example, when the movement trajectories, speeds and other factors of the target vehicle and the surrounding objects interact and may cause a collision, it is considered that there is a collision risk. In one implementation, the possibility and severity of the collision can be calculated based on factors such as the relative speed, distance and respective accelerations of the two vehicles. If the calculation result exceeds a certain threshold, it is considered that there is a collision risk.

[0038] Further, when it is determined that there is a collision risk, at least one collision plan and at least one avoidance plan are generated. Specifically, a collision plan refers to the collision response method adopted when the target vehicle cannot avoid a collision. When generating a collision plan, factors such as the physical characteristics of the vehicle itself and passenger safety need to be considered. An avoidance plan refers to the avoidance response method adopted by the target vehicle to avoid a collision. Exemplarily, the avoidance plans include emergency braking, emergency lane change, accelerating to avoid, etc. For example, the attitude of the target vehicle can be controlled to make the collision point as much as possible in the part with stronger vehicle structure, avoiding severe impact on key parts such as the cockpit; or adjusting the collision angle to make the collision energy more reasonably dispersed.

[0039] Step S120, calculate the collision cost of each collision plan and calculate the risk cost of each avoidance plan.

[0040] Collision cost refers to the quantitative representation of various losses and costs incurred when the target vehicle collides according to a certain collision scenario. Exemplarily, these losses include the degree of damage to the vehicle itself, the degree of injury that passengers in the vehicle may suffer, the damage caused to other surrounding objects, and the comprehensive costs such as the possible traffic interruption.

[0041] Risk cost refers to the quantitative representation of the risks and potential losses faced by the target vehicle when implementing a certain avoidance scenario. It can be understood that although the purpose of the avoidance scenario is to avoid collisions, there may be certain uncertainties and other risks during the implementation process. For example, the avoidance action may cause the vehicle to lose control, and there may be a new collision risk with other surrounding objects. These factors together constitute the risk cost of the avoidance scenario.

[0042] Optionally, before calculating the collision cost and risk cost, a corresponding cost calculation model is established. For the collision cost model, factors such as the physical structure, material, and position of key components of the vehicle are comprehensively considered to evaluate the vehicle damage cost; the protection effects of safety facilities such as human physiological characteristics and seat belts are used to evaluate the passenger injury cost; at the same time, historical traffic accident data can also be referred to analyze the impact degree of different types of collisions on traffic interruption, etc., so as to establish a mathematical model that can comprehensively reflect the collision losses.

[0043] For the risk cost model, factors such as the execution difficulty of the avoidance action, the complexity of the surrounding traffic environment, and traffic rule restrictions need to be analyzed. For example, when making an emergency lane change, the impact of factors such as the lane change distance, vehicle speed, and distance to the vehicle behind on the lane change safety needs to be considered, and then a model that can accurately measure the avoidance risk is established.

[0044] Then, using the vehicle's own sensors and the interaction information with the surrounding environment, the various parameters required for calculating the collision cost and risk cost are obtained. Next, the obtained parameters are substituted into the corresponding cost calculation model to calculate the collision cost of each collision scenario and the risk cost of each avoidance scenario respectively.

[0045] Step S130, based on the collision costs of each collision scenario and the risk costs of each avoidance scenario, determine the target driving strategy from at least one collision scenario and at least one avoidance scenario.

[0046] In this embodiment, the target driving strategy refers to the optimal driving strategy selected from the collision scenario and the avoidance scenario after comprehensively considering the collision cost and risk cost, aiming to balance safety and driving efficiency, so that the target vehicle can drive in the most reasonable way in the current traffic environment, which can not only minimize the losses caused by collisions to the greatest extent but also effectively avoid unnecessary risks.

[0047] In this embodiment, first, at least one collision avoidance plan and at least one risk avoidance plan are generated simultaneously, providing multiple possibilities for subsequent selection of a better driving strategy. Then, the collision costs of each collision avoidance plan and the risk costs of each risk avoidance plan are calculated, so as to intuitively display the potential losses and risks of different plans in a quantitative manner. Finally, based on cost comparison and comprehensive consideration, the target driving strategy is determined from the collision avoidance plans and the risk avoidance plans, fully considering the balance between safety and driving efficiency, and enabling flexible selection of the most appropriate driving method in a complex traffic environment.

[0048] Overall, this application realizes the forward-looking decision-making of the vehicle driving strategy, which not only helps to reduce the losses caused by collisions, but also can effectively avoid unnecessary risks, improves the vehicle's response ability when facing collision risks, enhances driving safety, and at the same time significantly improves the rationality and scientificity of the vehicle driving strategy, effectively reducing the probability of safety accidents and various possible losses.

[0049] Figure 2 The figure shows a schematic flowchart of generating at least one collision avoidance plan and at least one risk avoidance plan provided by an embodiment of this application. Figure 1 Based on the embodiment shown, Figure 2 the embodiment shown is extended, Figure 2 and the differences between the embodiment shown and Figure 1 the embodiment shown are mainly described below, and the same parts will not be elaborated.

[0050] As Figure 2 shown, in this embodiment, when there is a collision risk between the target vehicle and surrounding objects, at least one collision avoidance plan and at least one risk avoidance plan are generated, including the following steps.

[0051] Step S210, when it is determined that there is a collision risk between the target vehicle and surrounding objects, at least one risk avoidance plan is generated.

[0052] Optionally, once the collision risk is determined, at least one risk avoidance plan will be generated according to the current traffic conditions and vehicle driving state. For example, when the vehicle in front suddenly brakes sharply, and the target vehicle is close to the vehicle in front and has a high speed, emergency braking will be selected as the risk avoidance plan; if an obstacle suddenly appears on one side of the target vehicle and the vehicle behind is far away, emergency lane change will be selected to avoid the obstacle.

[0053] Specifically, in one implementation, the motion information of the target vehicle and the status data of surrounding objects are obtained; based on the motion information of the target vehicle and the status data of surrounding objects, a candidate risk avoidance sequence is generated; and each candidate risk avoidance sequence is subjected to a feasibility check, and the candidate risk avoidance sequence that passes the feasibility check is determined as the risk avoidance plan.

[0054] Exemplarily, the motion information includes information related to the driving state of the target vehicle, such as the speed, acceleration, driving direction, wheel angle, etc. The state data includes parameters describing the state of surrounding objects, such as their positions, speeds, sizes, types, etc.

[0055] Optionally, the target vehicle uses sensors such as cameras and wheel speed sensors to obtain its own motion information and the state data of surrounding objects in real time. Then, based on the collected motion information and state data, a candidate risk avoidance sequence containing different combined actions is generated using a pre-designed risk avoidance strategy algorithm. Among them, the candidate risk avoidance sequence includes combinations of one or more actions among steering, braking, and accelerating, such as steering first and then braking, or only accelerating. Finally, a series of rules and conditions are used to verify the candidate risk avoidance sequence to determine whether it can be safely and effectively executed in the current environment, avoiding collisions and not affecting the normal driving of the vehicle. It can be understood that only the candidate risk avoidance sequences that pass these verifications can be determined as the final risk avoidance solutions.

[0056] This solution first generates candidate risk avoidance sequences containing various action combinations, increasing the diversity and flexibility of the risk avoidance strategy, and enabling better response to complex and changeable traffic scenarios. Then, strict feasibility verification is performed on each candidate risk avoidance sequence to screen out safe, reliable, and feasible risk avoidance solutions, effectively reducing the risk of risk avoidance failure or causing secondary accidents, and enhancing the vehicle's response ability when facing collision risks.

[0057] Exemplarily, assume that the target vehicle is driving at a speed of 50 km / h on an urban road, and the vehicle in front suddenly brakes sharply. The distance between the target vehicle and the vehicle in front is relatively close, and there is a risk of collision. At this time, the target vehicle obtains information such as its own speed, acceleration, and the position and speed of the vehicle in front. Based on this information, multiple candidate risk avoidance sequences are generated, such as: Sequence 1: First perform emergency braking, and if the braking distance is insufficient, then change lanes to the left.

[0058] Sequence 2: Immediately change lanes to the right to avoid.

[0059] Sequence 3: Accelerate to overtake and avoid the vehicle in front that has suddenly stopped.

[0060] Then, verify Sequence 1: Calculate the braking distance according to the speed and braking performance of the target vehicle. If the calculated braking distance is greater than the actual distance from the vehicle in front, it is necessary to further determine whether the lane change action is feasible. Check whether there is enough space in the left lane for a lane change and no other vehicles blocking, and also consider whether the lane change complies with local traffic rules.

[0061] Verify Sequence 2: Check whether there are obstacles or vehicles in the right lane, whether the safety distance requirements for lane change are met, and whether the driving trajectory after lane change will cause collisions with other objects on the right.

[0062] Verify Sequence 3: Check whether accelerating to overtake complies with the traffic rules of the current road section (such as whether overtaking is allowed), whether there is enough space to pass safely after acceleration and whether it will collide with oncoming vehicles, etc.

[0063] Finally, only the candidate avoidance sequences that pass all verifications will be determined as avoidance plans. For example, if Sequence 1 meets the braking distance and lane change requirements in the current environment and the left lane change is compliant and safe, it will be determined as the avoidance plan and executed.

[0064] Step S220, generate corresponding collision plans based on at least one avoidance plan.

[0065] Optionally, for each avoidance plan, consider various failure situations that may occur during its actual execution. For example, the avoidance plan is to change lanes to the left, but other vehicles or obstacles may suddenly appear in the left lane, resulting in the inability to complete the lane change; or the avoidance plan is emergency braking, but the braking distance may be insufficient due to factors such as slippery roads, and the collision cannot be avoided. Then, based on the failure situations of the avoidance plan, determine the possible collision scenarios. For example, if the left lane change avoidance fails, there may be a side collision with a vehicle that suddenly appears on the left; if the emergency braking fails, there may be a rear-end collision with the vehicle in front. Finally, generate corresponding collision plans for the determined collision scenarios. In the collision plan, it is necessary to consider how to protect the safety of the vehicle occupants during the collision, minimize vehicle damage, and reduce the impact on other traffic participants.

[0066] More specifically, generating corresponding collision plans based on at least one avoidance plan includes: for each avoidance plan, predict the execution process of the avoidance plan; if the execution of the avoidance plan fails, determine the potential collision object and / or potential collision position after the execution of the avoidance plan; based on the potential collision object and / or potential collision position after the execution of the avoidance plan, generate the collision plan corresponding to the avoidance plan.

[0067] Execution failure means that the target vehicle cannot avoid colliding with surrounding objects. The potential collision object refers to the surrounding object that the target vehicle may collide with in the case of the failure of the avoidance plan execution. The potential collision position refers to the position where the target vehicle may collide in the case of the failure of the avoidance plan execution.

[0068] Optionally, the execution process of the risk avoidance plan is predicted through sensors and related algorithms. Taking an emergency lane change as an example, the sensors will monitor information such as the vehicle speed and steering wheel angle, and the related algorithms will predict the vehicle's lane change trajectory and the required time. If the execution of the risk avoidance plan fails, the potential collision object and location will be determined based on the prediction results. For example, when changing lanes, if a vehicle in the side and rear enters the blind spot, the potential collision object is that vehicle, and the potential collision location is at the side and rear of the target vehicle. Finally, a collision plan is generated based on the potential collision object and location, and the associated information of the collision plan includes the collision object and / or the collision location. Exemplarily, the collision plan includes adjusting the vehicle attitude and controlling the vehicle so that its side protection structure contacts the obstacle; or reducing the vehicle speed to reduce the collision speed and impact force.

[0069] This plan predicts the risk avoidance process, plans the collision response strategy in advance, effectively improves the safety of the vehicle in a complex traffic environment, and reduces accident losses.

[0070] Furthermore, the collision plan corresponding to this risk avoidance plan includes a collision attitude optimization instruction and / or a collision energy distribution strategy. Based on the potential collision object and / or potential collision location after the execution of this risk avoidance plan, a collision plan corresponding to this risk avoidance plan is generated, including: generating a collision attitude optimization instruction based on the type of the potential collision object and / or the potential collision location; and / or generating a collision energy distribution strategy based on the distribution of the potential collision location and the buffer zone of the target vehicle.

[0071] The collision attitude optimization instruction is used to indicate the optimization of the collision location and / or the collision angle, and is used to adjust the attitude of the vehicle during the collision. The purpose is to make the target vehicle contact the collision object in a relatively safe manner during the collision through reasonable attitude adjustment, so as to reduce the harm to the vehicle occupants and the damage to the vehicle. The collision energy distribution strategy includes a distribution rule for dispersing the impact energy generated by the collision to the buffer zone according to a preset ratio. The purpose is to reduce the impact force borne locally through reasonable energy distribution, thereby reducing damage.

[0072] Optionally, when generating the collision attitude optimization instruction, the potential collision object and location after the failure of the execution of the risk avoidance plan are determined, including information such as its type, size, location, etc., and its distribution around the target vehicle. Exemplarily, if it is a pedestrian, try to adjust the vehicle attitude so that the side of the vehicle body contacts the pedestrian, reducing the possibility of directly hitting the key parts of the pedestrian; if it is a vehicle, try to make the two vehicles contact at a relatively low speed and a small angle, avoiding a frontal rigid collision. Exemplarily, if the collision location is on the side of the vehicle, adjust the vehicle attitude so that the side protection structure contacts the collision object; if it is in the front, try to adjust the vehicle attitude so that the buffer zone such as the engine compartment can effectively absorb energy during a frontal collision.

[0073] Optionally, when generating a collision energy distribution strategy, determine the specific location where the collision occurs, such as the left front of the vehicle. It can be understood that the energy absorption capabilities and structural characteristics of different buffer zones are different, and then further combine with the distribution of buffer zones, such as the buffer zones of the vehicle are distributed at the front, rear, and sides of the vehicle. Finally, according to the collision position, reasonably distribute the impact energy to each buffer zone. For example, through vehicle structure design and material properties, part of the energy is guided to structures such as the longitudinal beams and cross beams of the vehicle frame, so that they jointly bear the impact force and avoid excessive local energy causing serious deformation or damage.

[0074] When the risk avoidance fails, this solution provides a final protection barrier for the vehicle by optimizing the collision attitude and energy distribution strategy, effectively reducing the damage caused by the collision, and protecting the personal safety and property safety of the vehicle occupants and surrounding traffic participants.

[0075] Generally speaking, in this embodiment, the solutions of step S210 and step S220 first generate at least one risk avoidance solution when there is a collision risk between the target vehicle and surrounding objects. On the basis of generating the risk avoidance solution, further generate the corresponding collision solution to prepare for the possible collision, forming a comprehensive risk response system.

[0076] Figure 3 The following shows a flowchart of a vehicle driving strategy determination method provided by another embodiment of the present application. Figure 1 Based on the embodiment shown above, Figure 3 the embodiment shown below is extended. Figure 3 The differences between the embodiment shown below and Figure 1 the embodiment shown above will be emphasized below, and the same parts will not be elaborated.

[0077] As Figure 3 shown, in this embodiment, before generating at least one collision solution and at least one risk avoidance solution when there is a collision risk between the target vehicle and surrounding objects, the following steps are further included.

[0078] Step S310, obtain the motion information of the target vehicle and the state data of surrounding objects.

[0079] Optionally, the motion information of the target vehicle includes at least one of the pose, speed, acceleration, kinematic parameters, dynamic parameters, etc. of the target vehicle. The state data of surrounding objects includes at least one of the position, speed, size, motion direction, etc. of surrounding objects.

[0080] Exemplarily, information about the target vehicle and surrounding objects can be collected through sensors (such as millimeter-wave radar, lidar, and cameras, etc.).

[0081] Step S320: Generate at least one predicted trajectory of the target vehicle within a future target time based on the motion information of the target vehicle.

[0082] A predicted trajectory refers to the path that the target vehicle is predicted to travel within a future target time based on its motion information, usually including position information at multiple time points.

[0083] Optionally, the obtained motion information (such as pose, speed, acceleration, etc.) is used as input, combined with the dynamic model or kinematic model of the target vehicle, to predict at least one predicted trajectory of the target vehicle within a future target time for subsequent collision risk assessment.

[0084] Step S330: If it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object, it is determined that there is a collision risk between the target vehicle and the surrounding object.

[0085] Specifically, the motion information includes at least one of pose, speed, acceleration, kinematic parameters, and dynamic parameters; the status data includes at least one of position, speed, size, and motion direction.

[0086] Overlap means that there is a situation where the predicted trajectory of the target vehicle and the future possible position of the surrounding object cross or cover in space.

[0087] Optionally, based on the status data of the surrounding object, predict the possible position range of the surrounding object within a future target time. Compare these position ranges with the predicted trajectory of the target vehicle to determine whether there is a spatial overlap. In one implementation, if there is an overlap, it is determined that there is a collision risk between the target vehicle and the surrounding object.

[0088] In this solution, by obtaining the motion information and status data of the target vehicle and surrounding objects in real time, generating the predicted trajectory of the target vehicle, and determining whether there is a collision risk with the surrounding objects, potential dangers can be identified in advance, effectively enhancing the active safety performance of the vehicle and improving the vehicle's risk perception and response capabilities in complex traffic environments.

[0089] In another implementation, if it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object, it includes: calculating the spatio-temporal overlap probability between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object; if the spatio-temporal overlap probability is greater than a preset probability threshold, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object.

[0090] Specifically, the spatio-temporal overlap probability refers to the likelihood of the predicted trajectory of the target vehicle overlapping with surrounding objects within a specific future time period and spatial region. The preset probability threshold is a pre-set probability value used to determine whether the spatio-temporal overlap probability reaches the standard sufficient to recognize the existence of a collision risk.

[0091] Optionally, according to the motion state of the surrounding objects, calculate their possible position distributions within the future target time, and combine with the predicted trajectory of the target vehicle to calculate the overlap probability of the two within the same time and the same spatial region. If the calculated spatio-temporal overlap probability is greater than the preset probability threshold, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects, that is, there is a collision risk.

[0092] Exemplarily, there is a pedestrian crossing the road in front of the target vehicle. According to the position, speed, and motion direction of the pedestrian, as well as the speed and predicted trajectory of the target vehicle, the calculated probability of spatio-temporal overlap between the target vehicle and the pedestrian in the crossing area within the next 2 seconds is 70%. If the preset probability threshold is 50%, it is determined that there is a collision risk.

[0093] In this solution, the calculation of the spatio-temporal overlap probability takes into account both time and space factors, making the risk assessment more comprehensive and accurate; secondly, the preset probability threshold provides a clear risk determination standard, improving the consistency and stability of the assessment; finally, this solution can identify potential collision risks earlier in complex traffic scenarios, providing more sufficient time for subsequent risk avoidance measures, thereby effectively reducing the probability of traffic accidents and improving the safety and reliability of driving.

[0094] Regarding step S120, in some embodiments, calculate the collision cost of each collision scenario, including: for each collision scenario, through at least one cost calculation function in the collision cost model, based on the costs corresponding to the collision object and / or collision position associated with this collision scenario, as well as the cost corresponding to the target vehicle, calculate the collision cost of this collision scenario.

[0095] Specifically, the cost of the collision object is determined based on the type of this collision object or the value of this collision object.

[0096] Optionally, take the costs corresponding to the collision object and / or collision position associated with the collision scenario, as well as the cost corresponding to the target vehicle as inputs. Then, through at least one cost calculation function in the collision cost model, comprehensively consider the above cost factors and calculate the total collision cost of this collision scenario.

[0097] Taking the cost corresponding to the collision position as an example, assume that the cost corresponding to the collision position of the collision object is 50,000 yuan; the cost of damage to the target vehicle itself is 30,000 yuan. Through the cost calculation function, these costs are combined to obtain a total collision cost of 80,000 yuan.

[0098] In this solution, the costs of the collision object and the collision position are incorporated into the calculation of the collision cost, enabling the accounting of the collision cost to more realistically reflect the actual possible losses. Secondly, the damage cost of the target vehicle itself is considered simultaneously, avoiding one-sidedness in cost assessment. This refined cost calculation method provides a more accurate basis for selecting the optimal driving strategy subsequently, and helps to make more reasonable decisions in complex traffic environments.

[0099] Regarding step S120, in some embodiments, the risk costs of each risk avoidance plan are calculated, including: for each risk avoidance plan, through at least one cost calculation function in the risk cost model, based on at least one of the type of risk avoidance action, the status data of surrounding objects, the motion information of the target vehicle, and the environmental parameters, the risk cost of this risk avoidance plan is calculated.

[0100] Specifically, the risk cost is the expected cost of the occurrence probability of risk events during the execution of the risk avoidance plan. More specifically, the occurrence probability of risk events refers to the probabilities of various risk events that may occur during the execution of the risk avoidance plan, such as the collision probability caused by the failure of risk avoidance, the probability of scraping with other vehicles, etc. The expected cost refers to the cost expectation value that may be generated during the execution of the risk avoidance plan calculated based on the occurrence probability of risk events.

[0101] Optionally, at least one of the type of risk avoidance action, the status data of surrounding objects, the motion information of the target vehicle, and the environmental parameters is used as input. Further, based on the input data, various risk events that may occur during the execution of the risk avoidance plan and their occurrence probabilities are analyzed. Then, through at least one cost calculation function in the risk cost model, combining the occurrence probability of risk events and the corresponding cost assessment, the risk cost of this risk avoidance plan is calculated. Among them, the type of risk avoidance action refers to the specific actions taken by the vehicle in the risk avoidance plan, such as emergency braking, emergency lane change, accelerating to avoid, etc. The environmental parameters refer to the environmental conditions where the target vehicle is located, such as road conditions, weather conditions, traffic flow, etc.

[0102] Exemplarily, assume that there is a vehicle that suddenly stops urgently in front of the target vehicle, and the generated risk avoidance plan is to change lanes to the left. Through the cost calculation function in the risk cost model, the following factors are considered: Type of risk avoidance action: Changing lanes to the left; Status data of surrounding objects: There is a vehicle with a relatively high speed in the left lane and the distance is relatively close; Motion information of the target vehicle: The current speed is 50 km / h, and the time required for lane change is 3 seconds; Environmental parameters: The road is slippery, the weather is light rain, and the visibility is low.

[0103] Finally, based on this data, the probability of a collision with the vehicle on the left during the execution of the risk avoidance plan is calculated to be 10%, as well as other possible risk events (such as the probability of the vehicle losing control being 5%) and their corresponding costs. By integrating the probabilities and costs of these risk events, the risk cost of this risk avoidance plan is calculated.

[0104] This plan calculates the risk cost of the risk avoidance plan by integrating multi-dimensional factors such as the type of risk avoidance actions, the status data of surrounding objects, the motion information of the target vehicle, and environmental parameters, using a risk cost model, making the risk cost calculation more scientific and targeted, providing a reliable basis for selecting the optimal risk avoidance strategy, and thus improving the decision-making rationality of the vehicle in a complex traffic environment.

[0105] Optionally, in some embodiments, the motion information of the target vehicle includes dynamic parameters, which include load and / or center of gravity height, and the environmental parameters where the target vehicle is located include road surface adhesion coefficient and / or visibility.

[0106] The road surface adhesion coefficient is a dimensionless parameter that measures the magnitude of the frictional force between the road surface and the tire. It reflects the adhesion ability between the vehicle tire and the road surface, and its value usually ranges between 0 and 1. It can be understood that the larger the adhesion coefficient, the greater the frictional force between the tire and the road surface, and the better the braking performance and handling stability of the vehicle. Visibility refers to the maximum distance at which a driver or the vehicle's perception system can clearly see an object or obstacle ahead under the current environmental conditions, usually measured in meters (m) or feet (ft).

[0107] Exemplarily, assume that the target vehicle is traveling on a slippery road surface with a low road surface adhesion coefficient and reduced visibility due to fog. The vehicle has a large load and a high center of gravity height. In the risk avoidance plan, the vehicle needs to make an emergency lane change. Due to the large load, the vehicle has a large inertia, and a greater centrifugal force is required during the lane change. The high center of gravity height may cause the vehicle to roll over more easily during the lane change. The low road surface adhesion coefficient reduces the frictional force between the tire and the road surface, reducing the stability of the lane change. The low visibility may delay the reaction time of the driver or the system, increasing the risk of risk avoidance failure. Considering these factors, the risk cost model will calculate a higher risk cost because these conditions increase the possibility of risk avoidance failure and may lead to more serious consequences.

[0108] This plan makes the calculation of the risk cost more accurate by considering various actual factors, which helps to make more reasonable risk avoidance decisions.

[0109] Figure 4 The following shows a schematic flowchart of the process for determining the target driving strategy provided by an embodiment of the present application. In Figure 1 Based on the embodiment shown, Figure 4 The following embodiment is extended, and the following will focus on describing Figure 4The differences between the illustrated embodiments and Figure 1 the illustrated embodiments will not be elaborated upon as the similarities are already known.

[0110] As Figure 4 shown, in this embodiment, based on the collision costs of each collision scenario and the risk costs of each avoidance scenario, a target driving strategy is determined from at least one collision scenario and at least one avoidance scenario, including the following steps.

[0111] Step S410: Compare the numerical values of the collision costs of the collision scenarios and the risk costs of the avoidance scenarios, and select the scenario with the smallest numerical value as the target driving strategy.

[0112] Specifically, obtain the collision costs of each collision scenario and the risk costs of each avoidance scenario. Then, compare the collision costs of the collision scenarios and the risk costs of the avoidance scenarios one by one, and select the scenario with the smallest cost value as the target driving strategy.

[0113] Exemplarily, in a complex traffic scenario, the target vehicle needs to choose between two scenarios: the collision cost of one collision scenario is 100,000 yuan, and the risk cost of another avoidance scenario is 80,000 yuan. By comparison, since 80,000 yuan is less than 100,000 yuan, this avoidance scenario is selected as the target driving strategy. This method based on cost comparison can intuitively select the optimal driving scenario in the current situation and balance safety and economy.

[0114] Step S420: Select a set of scenarios that meet the safety threshold from the collision scenarios and the avoidance scenarios, and select the scenario with the lowest cost from the set of scenarios as the target driving strategy.

[0115] The safety threshold refers to a pre-set minimum standard for measuring the safety of a scenario, and only scenarios that reach or exceed this standard will be considered.

[0116] Optionally, conduct a safety assessment on all collision scenarios and avoidance scenarios, screen out the scenarios that meet the safety threshold, and form a set of scenarios. Within the set of scenarios, select the scenario with the lowest cost (collision cost or risk cost) as the target driving strategy.

[0117] Exemplarily, the target vehicle faces multiple collision scenarios and avoidance scenarios. After the safety assessment, only Scenario C among the collision scenarios meets the safety threshold, and its collision cost is 120,000 yuan; among the avoidance scenarios, Scenarios A and B meet the safety threshold, and the risk costs are 80,000 yuan and 100,000 yuan respectively. At this time, the set of scenarios includes Scenarios C, A, and B. Compare the costs of these three scenarios, and finally select Scenario A with the lowest cost as the target driving strategy. This approach ensures that the most economically optimal driving scenario is selected on the basis of meeting safety requirements.

[0118] Figure 5 The figure shows a schematic flow chart of a vehicle mixed traffic control method provided by an embodiment of the present application. Exemplarily, as Figure 5 shown, the method includes the following steps.

[0119] Step S510: When there is a collision risk between a driverless vehicle and a manned vehicle, determine the target driving strategy corresponding to the driverless vehicle.

[0120] A driverless vehicle refers to a vehicle that can drive autonomously without manual driving. A manned vehicle refers to a vehicle controlled by a human driver. The target driving strategy is obtained based on the vehicle driving strategy determination method described in the foregoing embodiment.

[0121] Optionally, during the driving process of the driverless vehicle, the relative position, speed, etc. of the manned vehicle are monitored in real time to evaluate the collision risk. Once a risk is detected, the optimal target driving strategy is determined using the foregoing method.

[0122] Step S520: Control the driverless vehicle to drive based on the target driving strategy.

[0123] Optionally, the control system of the driverless vehicle generates specific control instructions according to the target driving strategy, such as adjusting the vehicle speed, steering, etc. During the driving process, the driverless vehicle continuously monitors the environmental changes and driving state, and dynamically adjusts the control instructions to ensure the effective execution of the driving strategy.

[0124] Exemplarily, the driverless vehicle detects that there is a manned vehicle suddenly braking sharply in front, and there is a collision risk. By evaluating the costs and risks of various possible driving strategies (such as emergency braking, lane changing, etc.), the optimal target driving strategy is determined to be emergency braking. Subsequently, the control system of the driverless vehicle executes the emergency braking instruction, and the vehicle decelerates smoothly, successfully avoiding a collision with the vehicle in front.

[0125] In this solution, when there is a collision risk between a driverless vehicle and a manned vehicle, the optimal driving strategy is determined using the foregoing method and the vehicle is controlled accordingly. On the one hand, it ensures that the driverless vehicle can make decisions quickly and accurately in the face of complex traffic conditions, reducing the probability of collision with manned vehicles; on the other hand, based on the comprehensive evaluation of costs and risks, the driving path and speed of the driverless vehicle are optimized, improving the driving efficiency.

[0126] As described above in combination with Figures 1 to 5 , the method embodiments of the present application have been described in detail. Next, in combination with Figure 6 and Figure 7 , the device embodiments of the present application will be described in detail. It should be understood that the descriptions of the method embodiments and the device embodiments correspond to each other. Therefore, for the parts not described in detail, reference can be made to the foregoing method embodiments.

[0127] Figure 6 The following is a schematic structural diagram of a vehicle driving strategy determination device provided by an embodiment of the present application. As Figure 6 shown, the vehicle driving strategy determination device 60 provided by the embodiment of the present application includes: A generation module 610, configured to generate at least one collision plan and at least one avoidance plan when there is a collision risk between the target vehicle and surrounding objects. The collision plan refers to the collision response method adopted when the target vehicle cannot avoid a collision, and the avoidance plan refers to the avoidance response method adopted by the target vehicle to avoid a collision; A calculation module 620, configured to calculate the collision cost of each collision plan and calculate the risk cost of each avoidance plan; A first determination module 630, configured to determine a target driving strategy from at least one collision plan and at least one avoidance plan based on the collision cost of each collision plan and the risk cost of each avoidance plan.

[0128] In an embodiment of the present application, the generation module 610 is further configured to generate at least one avoidance plan when it is determined that there is a collision risk between the target vehicle and surrounding objects; and generate a corresponding collision plan based on the at least one avoidance plan.

[0129] In an embodiment of the present application, the generation module 610 is further configured to obtain the motion information of the target vehicle and the state data of surrounding objects; generate a candidate avoidance sequence based on the motion information of the target vehicle and the state data of surrounding objects, where the candidate avoidance sequence includes a combination of one or more actions among steering, braking, and accelerating; perform a feasibility check on each candidate avoidance sequence, and determine the candidate avoidance sequence that passes the feasibility check as an avoidance plan.

[0130] In an embodiment of the present application, the generation module 610 is further configured to, for each avoidance plan, predict the execution process of the avoidance plan; if the execution of the avoidance plan fails, determine the potential collision object and / or potential collision position after the execution of the avoidance plan, where the execution failure means that the target vehicle cannot avoid a collision with surrounding objects; generate a collision plan corresponding to the avoidance plan based on the potential collision object and / or potential collision position after the execution of the avoidance plan, and the associated information of the collision plan includes the collision object and / or the collision position.

[0131] In an embodiment of the present application, the collision plan corresponding to the risk avoidance plan includes a collision attitude optimization instruction and / or a collision energy distribution strategy; the generation module 610 is further configured to generate a collision attitude optimization instruction based on the type of the potential collision object and / or the potential collision position, and the collision attitude optimization instruction is used to indicate the optimization of the collision position and / or the collision angle; and / or, generate a collision energy distribution strategy based on the distribution of the potential collision position and the buffer area of the target vehicle, and the collision energy distribution strategy includes a distribution rule for dispersing the impact energy generated by the collision to the buffer area according to a preset ratio.

[0132] In an embodiment of the present application, the generation module 610 is further configured to obtain the motion information of the target vehicle and the state data of the surrounding objects; generate at least one predicted trajectory of the target vehicle within a future target time based on the motion information of the target vehicle; if it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects based on the state data of the surrounding objects, it is determined that there is a collision risk between the target vehicle and the surrounding object; wherein, the motion information includes at least one of pose, speed, acceleration, kinematic parameters, and dynamic parameters; the state data includes at least one of position, speed, size, and motion direction.

[0133] In an embodiment of the present application, the generation module 610 is further configured to calculate the spatio-temporal overlap probability between the predicted trajectory of the target vehicle and the surrounding objects based on the state data of the surrounding objects; if the spatio-temporal overlap probability is greater than a preset probability threshold, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding objects.

[0134] In an embodiment of the present application, the calculation module 620 is further configured to, for each collision plan, calculate the collision cost of the collision plan through at least one cost calculation function in the collision cost model, based on the costs corresponding to the collision object and / or the collision position associated with the collision plan, and the cost corresponding to the target vehicle; wherein, the cost of the collision object is determined based on the type of the collision object or the value of the collision object.

[0135] In an embodiment of the present application, the calculation module 620 is further configured to, for each risk avoidance plan, calculate the risk cost of the risk avoidance plan through at least one cost calculation function in the risk cost model, based on at least one of the type of the risk avoidance action, the state data of the surrounding objects, the motion information of the target vehicle, and the environmental parameters, wherein the risk cost is the expected cost of the occurrence probability of a risk event during the execution of the risk avoidance plan.

[0136] In an embodiment of the present application, the motion information of the target vehicle includes dynamic parameters, and the dynamic parameters include load and / or center of gravity height, and the environmental parameters of the target vehicle include road surface adhesion coefficient and / or visibility.

[0137] In an embodiment of the present application, the first determination module 630 is further configured to compare the numerical values of the collision cost of the collision plan and the risk cost of the avoidance plan, and select the plan with the smallest numerical value as the target driving strategy; alternatively, select a set of plans that meet the safety threshold from the collision plan and the avoidance plan, and select the plan with the lowest cost from the set of plans as the target driving strategy.

[0138] Figure 7 The following shows a schematic structural diagram of a vehicle mixed-traffic control device provided by an embodiment of the present application. As Figure 7 shown, the vehicle mixed-traffic control device 70 provided by the embodiment of the present application includes: A second determination module 710, configured to determine a target driving strategy corresponding to the driverless vehicle when there is a collision risk between the driverless vehicle and the manned vehicle; A control module 720, configured to control the driving of the driverless vehicle based on the target driving strategy.

[0139] Next, reference is made to Figure 8 to describe the driverless vehicle according to the embodiment of the present application. Figure 8 The following shows a schematic structural diagram of a driverless vehicle provided by an exemplary embodiment of the present application.

[0140] As Figure 8 shown, the driverless vehicle 80 includes one or more processors 801 and a memory 802.

[0141] The processor 801 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the driverless vehicle 80 to perform desired functions.

[0142] The memory 802 may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage media, and the processor 801 may run the program instructions to implement the methods of the various embodiments of the present application described above and / or other desired functions. Various contents such as collision plans, avoidance plans, collision costs, and avoidance costs may also be stored in the computer-readable storage media.

[0143] In one example, the driverless vehicle 80 may further include: an input device 803 and an output device 804, and these components are interconnected through a bus system and / or other forms of connection mechanisms (not shown).

[0144] The input device 803 may include, for example, a keyboard, a mouse, and the like.

[0145] The output device 804 may output various information to the outside, including a collision plan, an avoidance plan, a collision cost, an avoidance cost, etc. The output device 804 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, and the like.

[0146] Of course, for simplicity, Figure 8 only some of the components related to the present application in the driverless vehicle 80 are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, according to specific application scenarios, the driverless vehicle 80 may further include any other appropriate components.

[0147] In addition to the above methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, and when the computer program instructions are run by a processor, the processor is caused to execute the steps in the methods according to various embodiments of the present application described above in this specification.

[0148] The computer program product may be written in any combination of one or more programming languages for program code to perform the operations of the embodiments of the present application. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may be executed entirely on a user computing device, partially on the user device, executed as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0149] In addition, an embodiment of the present application may also be a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are run by a processor, the processor is caused to execute the steps in the methods according to various embodiments of the present application described above in this specification.

[0150] The computer-readable storage medium may adopt any combination of one or more readable media. The readable media may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0151] The basic principles of the present application have been described above in conjunction with specific embodiments. However, it should be noted that the advantages, benefits, effects, etc. mentioned in the present application are only examples and not limitations. It cannot be considered that these advantages, benefits, effects, etc. are essential for each embodiment of the present application. Additionally, the above-disclosed specific details are only for illustrative and facilitating understanding purposes and are not limitations. The above details do not limit the present application to necessarily adopt the above specific details for implementation.

[0152] The block diagrams of the devices, apparatuses, equipment, and systems involved in the present application are only illustrative examples and do not intend to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open-ended words, meaning "including but not limited to", and can be used interchangeably with each other. The word "or" and "and" used herein refer to the word "and / or" and can be used interchangeably with it, unless the context clearly indicates otherwise. The word "such as" used herein refers to the phrase "such as but not limited to" and can be used interchangeably with it.

[0153] It should also be noted that in the devices, equipment, and methods of the present application, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent solutions of the present application.

[0154] The above description of the disclosed aspects enables any person skilled in the art to make or use the present application. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

[0155] The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit embodiments of the present application to the form disclosed herein. Although several example aspects and embodiments have been discussed above, those skilled in the art will recognize some of their variations, modifications, alterations, additions, and subcombinations.

Claims

1. A method for determining a vehicle driving strategy, characterized in that Including: When there is a risk of collision between the target vehicle and surrounding objects, generating at least one collision plan and at least one avoidance plan, where the collision plan refers to the collision response method adopted when the target vehicle cannot avoid a collision, and the avoidance plan refers to the avoidance response method adopted by the target vehicle to avoid a collision; Calculating the collision cost of each collision plan and calculating the risk cost of each avoidance plan; Based on the collision cost of each collision plan and the risk cost of each avoidance plan, determining a target driving strategy from the at least one collision plan and the at least one avoidance plan.

2. The vehicle driving strategy determination method according to claim 1, characterized in that, The generating at least one collision plan and at least one avoidance plan when there is a risk of collision between the target vehicle and surrounding objects includes: When it is determined that there is a risk of collision between the target vehicle and the surrounding objects, generating the at least one avoidance plan; Based on the at least one avoidance plan, generating a corresponding collision plan.

3. The method for determining a vehicle driving strategy according to claim 2, wherein The generating at least one avoidance plan includes: Obtaining the motion information of the target vehicle and the status data of the surrounding objects; Based on the motion information of the target vehicle and the status data of the surrounding objects, generating a candidate avoidance sequence, where the candidate avoidance sequence includes a combination of one or more actions among steering, braking, and acceleration; Performing a feasibility check on each candidate avoidance sequence, and determining the candidate avoidance sequence that passes the feasibility check as the avoidance plan.

4. The method for determining a vehicle driving strategy according to claim 2, characterized in that, The generating a corresponding collision plan based on the at least one avoidance plan includes: For each avoidance plan, predicting the execution process of this avoidance plan; If the execution of this avoidance plan fails, determining the potential collision object and / or potential collision position after the execution of this avoidance plan, where the execution failure means that the target vehicle cannot avoid a collision with the surrounding objects; Based on the potential collision object and / or potential collision position after the execution of this avoidance plan, generating a collision plan corresponding to this avoidance plan, and the associated information of the collision plan includes the collision object and / or the collision position.

5. The vehicle driving strategy determination method according to claim 4, characterized in that The collision plan corresponding to this avoidance plan includes a collision attitude optimization instruction and / or a collision energy distribution strategy; The generating a collision plan corresponding to this avoidance plan based on the potential collision object and / or potential collision position after the execution of this avoidance plan includes: Based on the type of the potential collision object and / or the potential collision position, generating the collision attitude optimization instruction, where the collision attitude optimization instruction is used to indicate collision position optimization and / or collision angle optimization; And / or, based on the distribution of the potential collision position and the buffer zone of the target vehicle, generating a collision energy distribution strategy, where the collision energy distribution strategy includes a distribution rule for dispersing the impact energy generated by the collision to the buffer zone according to a preset ratio.

6. The method for determining a vehicle driving strategy according to any one of claims 1 to 5, characterized in that, Before generating at least one collision plan and at least one avoidance plan when there is a risk of collision between the target vehicle and surrounding objects, it further includes: Obtaining the motion information of the target vehicle and the status data of the surrounding objects; Based on the motion information of the target vehicle, generating at least one predicted trajectory of the target vehicle at a future target time; If it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object, it is determined that there is a collision risk between the target vehicle and the surrounding object; Wherein, the motion information includes at least one of pose, speed, acceleration, kinematic parameters, and dynamic parameters; the status data includes at least one of position, speed, size, and motion direction.

7. The method for determining a vehicle driving strategy according to claim 6, wherein The step of, if it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object, includes: Calculating the spatio-temporal overlap probability between the predicted trajectory of the target vehicle and the surrounding object based on the status data of the surrounding object; If the spatio-temporal overlap probability is greater than a preset probability threshold, it is determined that there is an overlap between the predicted trajectory of the target vehicle and the surrounding object.

8. The method for determining a vehicle driving strategy according to any one of claims 1 to 5, characterized in that The step of calculating the collision cost of each collision scenario includes: For each collision scenario, calculating the collision cost of the collision scenario through at least one cost calculation function in the collision cost model, based on the costs corresponding to the collision object and / or collision position associated with the collision scenario, and the cost corresponding to the target vehicle; Wherein, the cost of the collision object is determined based on the type of the collision object or the value of the collision object.

9. The method for determining a vehicle driving strategy according to any one of claims 1 to 5, characterized in that, The step of calculating the risk cost of each avoidance scenario includes: For each avoidance scenario, calculating the risk cost of the avoidance scenario through at least one cost calculation function in the risk cost model, based on at least one of the type of the avoidance action, the status data of the surrounding object, the motion information of the target vehicle, and the environmental parameters of the location, wherein the risk cost is the expected cost of the occurrence probability of a risk event during the execution of the avoidance scenario.

10. The vehicle driving strategy determination method according to claim 9, characterized in that, The motion information of the target vehicle includes dynamic parameters, and the dynamic parameters include load and / or center of gravity height, and the environmental parameters of the location where the target vehicle is located include road surface adhesion coefficient and / or visibility.

11. The method for determining a vehicle driving strategy according to any one of claims 1 to 5, characterized in that, The step of determining the target driving strategy from the at least one collision scenario and the at least one avoidance scenario based on the collision costs of the respective collision scenarios and the risk costs of the respective avoidance scenarios includes: Comparing the numerical values of the collision costs of the collision scenarios and the risk costs of the avoidance scenarios, and selecting the scenario with the smallest numerical value as the target driving strategy; Alternatively, selecting a set of scenarios that meet the safety threshold from the collision scenarios and the avoidance scenarios, and selecting the scenario with the lowest cost from the set of scenarios as the target driving strategy.

12. A vehicle mixed traffic control method, characterized in that, including: In the case where there is a collision risk between the driverless vehicle and the manned vehicle, determining the target driving strategy corresponding to the driverless vehicle, wherein the target driving strategy is obtained based on the method according to any one of claims 1 to 11; Controlling the driving of the driverless vehicle based on the target driving strategy.

13. An autonomous vehicle, characterized in that, including: a processor; a memory for storing instructions executable by the processor; The processor is configured to execute the vehicle driving strategy determination method according to any one of claims 1 to 11, or the vehicle mixed driving control method according to claim 12.

Images