Vehicle control method, vehicle, storage medium, and electronic device

Abstract

Embodiments of the present application provide a vehicle control method, a vehicle, a storage medium and an electronic device. The method comprises: obtaining a longitudinal distance between the vehicle and a target obstacle; in response to the longitudinal distance being less than a safe braking distance, determining a target deceleration of the vehicle according to a real-time acceleration of the vehicle and a preset transition duration, wherein the preset transition duration is used to represent a duration required for the vehicle to adjust from the real-time acceleration to the target deceleration; and controlling the vehicle to avoid the target obstacle based on the target deceleration. The present application solves the technical problem of poor emergency obstacle avoidance effect of an auxiliary driving system in the related art.

Description

Technical Field

[0001] This application relates to the field of vehicle active safety technology, and more specifically, to a vehicle control method, a vehicle, a storage medium, and an electronic device. Background Technology

[0002] The emergency obstacle avoidance function of an assisted driving system is used to guide the vehicle safely away from a dangerous path when a collision risk is detected ahead and conventional braking cannot avoid an accident. Although vehicles with assisted driving functions are becoming increasingly common, the scope of application scenarios for assisted driving functions is constantly expanding, and more and more drivers are using combined assisted driving functions, existing combined assisted driving functions often lack sufficient response to some emergency situations. For example, they often cannot effectively avoid collisions when faced with a rear-end collision scenario where the vehicle in front suddenly slows down, or when a pedestrian suddenly appears out of the way from behind.

[0003] There is currently no good solution to the above problems. Summary of the Invention

[0004] This application provides a vehicle control method, a vehicle, a storage medium, and an electronic device to at least solve the technical problem of poor emergency obstacle avoidance performance of driver assistance systems in related technologies.

[0005] According to one aspect of the embodiments of this application, a vehicle control method is provided, comprising: acquiring the longitudinal distance between a vehicle and a target obstacle; in response to the longitudinal distance being less than a safe braking distance, determining a target deceleration of the vehicle based on the vehicle's real-time acceleration and a preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration; and controlling the vehicle to avoid the target obstacle based on the target deceleration.

[0006] Furthermore, controlling the vehicle to avoid a target obstacle based on the target deceleration includes: in response to the target deceleration being greater than a preset deceleration threshold and the existence of steering avoidance space, controlling the vehicle to make an emergency turn to avoid the target obstacle, wherein the preset deceleration threshold is used to represent the upper limit of deceleration within a safe range.

[0007] Furthermore, controlling the vehicle to avoid a target obstacle based on the target deceleration includes: in response to the target deceleration being less than or equal to a preset deceleration threshold, controlling the vehicle to perform emergency braking to avoid the target obstacle.

[0008] Furthermore, the vehicle control method also includes: when controlling the vehicle to perform emergency braking, in response to the risk of collision between the vehicle and the target obstacle, determining a steering collision avoidance point, wherein the steering collision avoidance point is used to indicate the starting position of the steering to safely avoid the target obstacle; and controlling the vehicle to perform an emergency steering at the steering collision avoidance point to avoid the target obstacle.

[0009] Furthermore, the target deceleration of the vehicle is determined based on the vehicle's real-time acceleration and preset transition time, including: determining multiple acceleration change rates based on the real-time acceleration and preset transition time; determining multiple decelerations based on the multiple acceleration change rates and preset transition time; and determining the deceleration whose absolute value among the multiple decelerations meets a first preset standard as the target deceleration.

[0010] Furthermore, the vehicle control method also includes: acquiring multiple obstacles around the vehicle; filtering the multiple obstacles according to the positional relationship between the multiple obstacles and the vehicle to obtain at least one initial obstacle; and determining a target obstacle that meets the second preset standard by performing lateral risk verification and longitudinal risk verification on the at least one initial obstacle.

[0011] Furthermore, the vehicle control method also includes: controlling the vehicle to make an emergency turn based on a planned path, wherein the planned path is used to avoid a target obstacle; and disengaging from controlling the vehicle to make an emergency turn in response to the presence of a preset obstacle on the planned path.

[0012] According to another aspect of the embodiments of this application, a vehicle control device is also provided, including: an acquisition module for acquiring the longitudinal distance between the vehicle and a target obstacle; a determination module for determining a target deceleration of the vehicle based on the vehicle's real-time acceleration and a preset transition time in response to the longitudinal distance being less than a safe braking distance, wherein the preset transition time is used to represent the time required for the vehicle to adjust from real-time acceleration to target deceleration; and a control module for controlling the vehicle to avoid the target obstacle based on the target deceleration.

[0013] Furthermore, the control module is also used to control the vehicle to make an emergency turn to avoid the target obstacle in response to the target deceleration being greater than a preset deceleration threshold and the existence of steering avoidance space, wherein the preset deceleration threshold is used to represent the upper limit of deceleration within a safe range.

[0014] Furthermore, the control module is also used to control the vehicle to perform emergency braking in response to the target deceleration being less than or equal to a preset deceleration threshold, so as to avoid the target obstacle.

[0015] Furthermore, the control module is also used to determine a steering collision avoidance point in response to a risk of collision between the vehicle and a target obstacle when the vehicle is under emergency braking. The steering collision avoidance point is used to indicate the starting position of the steering to safely avoid the target obstacle. At the steering collision avoidance point, the vehicle is controlled to make an emergency steering maneuver to avoid the target obstacle.

[0016] Furthermore, the determining module is also used to determine multiple acceleration change rates based on real-time acceleration and preset transition time; determine multiple decelerations based on multiple acceleration change rates and preset transition time; and determine the deceleration whose absolute value among the multiple decelerations meets a first preset standard as the target deceleration.

[0017] Furthermore, the acquisition module is also used to acquire multiple obstacles around the vehicle; to filter the multiple obstacles according to the positional relationship between the multiple obstacles and the vehicle to obtain at least one initial obstacle; and to determine the target obstacle that meets the second preset standard by performing lateral risk verification and longitudinal risk verification on at least one initial obstacle.

[0018] Furthermore, the control module is also used to control the vehicle to make emergency turns based on the planned path, wherein the planned path is used to avoid target obstacles; in response to the presence of a preset obstacle on the planned path, the control module exits control of the vehicle to make emergency turns.

[0019] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the executable program, wherein the executable program executes the vehicle control method described in any of the above embodiments when running on the processor.

[0020] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to execute the vehicle control method described in any of the above when it is run on a computer or processor.

[0021] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to perform the vehicle control method of any of the above.

[0022] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the vehicle control methods in various embodiments of this application.

[0023] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the vehicle control method in various embodiments of this application.

[0024] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the vehicle control methods in various embodiments of this application.

[0025] In this embodiment, the longitudinal distance between the vehicle and the target obstacle is first obtained. Then, in response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration and the preset transition time. The preset transition time represents the time required for the vehicle to adjust from the real-time acceleration to the target deceleration. Finally, the vehicle is controlled to avoid the target obstacle based on the target deceleration. By co-calculating the lateral avoidance path and the longitudinal braking demand, the feasibility of steering avoidance and the braking efficiency are dynamically balanced. This achieves the purpose of triggering the automatic emergency steering function when the automatic emergency braking is ineffective. Thus, the technical effect of minimizing lateral displacement to achieve the best collision avoidance effect is achieved while still being able to avoid a collision. This solves the technical problem of poor emergency obstacle avoidance effect of the assisted driving system in the related art. Attached Figure Description

[0026] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0027] Figure 1 This is a flowchart of a vehicle control method according to an embodiment of this application;

[0028] Figure 2 This is a schematic diagram illustrating an emergency steering scenario triggered in a high-speed situation according to an embodiment of this application;

[0029] Figure 3 This is a schematic diagram illustrating an emergency steering scenario triggered in a low-speed environment according to an embodiment of this application.

[0030] Figure 4 This is a schematic diagram of an optional automatic emergency steering process according to an embodiment of this application;

[0031] Figure 5 This is a schematic diagram of an optional vehicle acceleration curve according to an embodiment of this application;

[0032] Figure 6 This is a schematic diagram of a vehicle control device according to an embodiment of this application. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0034] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0035] According to an embodiment of this application, an embodiment of a vehicle control method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0036] This embodiment provides a vehicle control method. Figure 1 This is a flowchart of a vehicle control method according to an embodiment of this application, such as... Figure 1 As shown, the process includes the following steps:

[0037] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0038] In this embodiment of the application, the vehicle refers to an intelligent connected vehicle equipped with an advanced driver assistance system and automatic emergency steering (AES) function, and equipped with a multi-sensor fusion system (including millimeter-wave radar, forward-facing camera, ultrasonic radar and high-precision map positioning module), which has the ability to perceive the surrounding environment in real time, calculate its own motion state and perform active intervention, and is the subject and object of this technical solution.

[0039] Target obstacles refer to dynamic or static objects identified and classified by the perception fusion module that are located in front of or adjacent to the vehicle's path and may pose a collision risk. These include, but are not limited to, vehicles slowing down or stationary ahead, pedestrians or cyclists suddenly crossing the road, road construction cones, roadside guardrails, barriers, and slow-moving non-motorized vehicles. The target obstacles are prioritized by the driver assistance system based on their type, location, speed, acceleration, confidence level, and predicted trajectory, serving as the core objects for collision avoidance decisions.

[0040] Longitudinal distance refers to the shortest straight-line distance measured along the vehicle's current direction of travel (i.e., longitudinal axis) from the vehicle's center of gravity to the front end of the target obstacle or its corresponding projection point. It is used to assess the Time-to-Collision (TTC) risk and determine whether to trigger the obstacle avoidance logic corresponding to the Forward Collision Warning (FCW). It is a key input parameter for deciding whether Automatic Emergency Braking (AEB) or Automatic Emergency Braking (AES) functions should intervene. For example, longitudinal distance is not a static coordinate difference, but a real-time dynamic quantity that updates dynamically with the relative speed, acceleration, and perception delay between the vehicle and the target. The calculation of longitudinal distance requires integrating the vehicle's own heading angle, the target's motion vector, sensor coordinate system transformation, and the predicted trajectory model, directly determining whether the driver assistance system has sufficient braking or steering space to complete effective collision avoidance.

[0041] Obtaining the longitudinal distance between the vehicle and the target obstacle can be understood as the real-time measurement of the straight-line distance between the vehicle and the nearest point of the target obstacle in the direction of the vehicle's movement, based on the vehicle's perception system (such as radar, camera, or fusion algorithm). This longitudinal distance is calculated by projecting it along the vehicle's longitudinal axis and is used to characterize the approaching situation between the vehicle and the target obstacle. It is the core input parameter for assessing the collision risk time and determining whether to trigger the automatic emergency braking or automatic emergency steering function.

[0042] It can be seen that by accurately obtaining the longitudinal distance between the vehicle and the target obstacle, key real-time data is provided for collision risk assessment, which effectively supports the accurate calculation of collision risk time, enabling the AEB and AES functions to intervene precisely at the right time, avoiding premature false triggering or late failure, significantly improving the response reliability and safety of high-speed collision avoidance, and enhancing the system's actual risk avoidance capabilities in scenarios such as sudden pedestrian appearances and sudden braking by the vehicle in front.

[0043] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0044] In this embodiment, the safe braking distance refers to the longest longitudinal braking distance that the vehicle can sustain from its current state when applying maximum effective braking to avoid a collision with a target obstacle, given the current vehicle speed, road adhesion conditions, and braking system response capability. The safe braking distance comprehensively considers perception delay, control system response delay, the hysteresis characteristics of the braking system's pressure build-up, and occupant comfort constraints. It is the critical threshold for the driver assistance system to determine whether braking alone can avoid a collision. If the actual longitudinal distance is less than the safe braking distance, it indicates that braking alone is insufficient to achieve safe obstacle avoidance, and a steering assist strategy must be triggered.

[0045] Real-time acceleration refers to the actual longitudinal acceleration experienced by a vehicle at the current moment. It is estimated jointly by the onboard inertial measurement unit or wheel speed sensors and the vehicle's dynamics model, reflecting the vehicle's current state of motion. Examples include the small positive acceleration a vehicle may be experiencing while cruising, the natural deceleration caused by an incline, or the dynamic changes resulting from the driver's slight application of the accelerator or brake. Real-time acceleration is the basis for driver assistance systems to determine whether the vehicle is in a controllable deceleration state, used to calculate the energy change and time required to transition from the current state to the target deceleration.

[0046] The preset transition time is a control parameter pre-calibrated and stored by the system, describing the length of time required for the vehicle to smoothly transition from the current real-time acceleration to the target deceleration. The preset transition time is not fixed but dynamically configured based on vehicle mass, braking system response characteristics, suspension stiffness, tire grip, and comfort strategies. For example, during high-speed cruising, to avoid discomfort to occupants due to sudden deceleration, the preset transition time may be set to 1.0~1.5 seconds to achieve a linear, gradual increase in deceleration; while in emergency collision avoidance scenarios, this time may be shortened to 0.3~0.6 seconds to approach the physical limits of the braking system and achieve the shortest stopping distance—this is not limited here. The core function of the preset transition time is to achieve an engineering balance between "control smoothness" and "response timeliness," preventing vehicle pitch, tire slippage, or occupant loss of control caused by sudden acceleration changes.

[0047] The target deceleration is the optimal longitudinal deceleration command value required to completely avoid a collision within the remaining distance, dynamically calculated by the driver assistance system through a control algorithm based on the comparison between the longitudinal distance and the safe braking distance, combined with the vehicle dynamics model, road conditions, and the upper limit of braking capacity. The target deceleration is typically greater than the current real-time acceleration and is limited by the maximum deceleration that the braking system can request (e.g., 0.8g~1.2g), the tire's limit coefficient of adhesion, and the intervention threshold of the electronic stability control system. The target deceleration does not simply pursue the maximum value, but rather aims to optimize the "minimum acceptable deceleration," minimizing disturbances to occupant comfort and vehicle stability while ensuring safety. The calculation process comprehensively considers obstacle motion trends, vehicle inertia, energy recovery efficiency, and brake fade characteristics.

[0048] Responding to a longitudinal distance less than the safe braking distance, determining the vehicle's target deceleration based on the vehicle's real-time acceleration and a preset transition duration can be understood as follows: when the driver assistance system determines that the longitudinal distance between the current vehicle and the obstacle ahead is insufficient for safe avoidance relying solely on braking, the controller no longer simply outputs the maximum braking deceleration command. Instead, it performs dynamic compensation calculations based on the vehicle's actual motion state. At this point, the driver assistance system needs to comprehensively consider the vehicle's physical inertia, braking system response characteristics, and occupant comfort boundaries, and plan a continuous curve for a smooth transition from the current acceleration to the target deceleration based on the preset transition duration. The preset transition duration can be dynamically adjusted based on vehicle weight, tire-road adhesion coefficient, brake hydraulic build-up rate, electronic stability system response capability, and driving mode (such as Eco, Comfort, Sport). For example, on highways, to avoid rear-end collisions or occupants lurching forward due to sudden braking, the system may choose a 1.2-second transition duration, allowing deceleration to gradually increase linearly. In urban scenarios with sudden pedestrian appearances, the system may compress the transition duration to less than 0.4 seconds to approach the physical limits of the braking system and strive for the shortest braking distance. The target deceleration is the minimum necessary deceleration value obtained by inversely solving the longitudinal dynamics model. This ensures that within the remaining longitudinal distance, even considering the movement trend of the target obstacle, perception errors, and system delays, the minimum distance between the vehicle and the obstacle is always greater than zero. This achieves the intelligent braking decision logic of "safety first, comfort in balance, and smooth control," significantly improving the system's collision avoidance success rate in high-risk scenarios while reducing the risk of loss of control due to drastic changes in longitudinal acceleration, occupant discomfort, or secondary accidents caused by insufficient reaction from following vehicles.

[0049] As can be seen, by combining real-time acceleration with a preset transition time to dynamically calculate the target deceleration, a smooth and gradual braking intervention is achieved, avoiding occupant discomfort, vehicle pitch, or the risk of rear-end collisions caused by abrupt braking. Compared to traditional fixed deceleration strategies, the technical solution of this application is more closely aligned with the actual motion state of the vehicle, significantly enhancing system reliability, especially at high speeds and on low-friction surfaces. Simultaneously, this step provides a stable longitudinal foundation for subsequent AES steering intervention, ensuring safer and more natural lateral and longitudinal coordinated collision avoidance, balancing safety, comfort, and control coordination.

[0050] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0051] In this embodiment, controlling vehicle obstacle avoidance based on target deceleration can be understood as follows: the driver assistance system does not simply execute "maximum braking" or "fixed deceleration" commands, but rather implements refined longitudinal control based on a calculated target deceleration value that matches the current vehicle speed, obstacle dynamics, longitudinal distance, and vehicle dynamics. This process fully considers the response delay of the braking system, tire adhesion, and the synergistic characteristics of electric and hydraulic braking to ensure smooth, continuous, and executable deceleration commands, avoiding vehicle instability or tire slippage due to sudden command changes. By precisely matching the target deceleration, the driver assistance system achieves optimal deceleration within a limited distance, ensuring a safe distance from obstacles while minimizing unnecessary over-braking, reducing energy loss and braking system wear. Simultaneously, this control strategy provides a stable motion reference for lateral avoidance (AES), enabling the vehicle to complete steering and collision avoidance maneuvers more accurately and safely while maintaining longitudinal controllability. This significantly improves overall safety performance and system synergy in complex collision avoidance scenarios, truly achieving a closed-loop active safety control.

[0052] As can be seen, precise vehicle avoidance based on target deceleration ensures efficient deceleration within a limited longitudinal distance, maximizing the avoidance of collision risks. This control strategy avoids vehicle instability and occupant discomfort caused by sudden braking, improving the smoothness and controllability of the braking process. Simultaneously, it provides a stable longitudinal foundation for the AES steering function, making lateral and longitudinal collision avoidance more precise and coordinated, reducing false triggering and excessive intervention. By dynamically adapting deceleration commands, the system optimizes energy consumption and braking system load while ensuring safety, enhancing the intelligence level of the vehicle's active safety systems.

[0053] Through the above steps, the longitudinal distance between the vehicle and the target obstacle is first obtained. Then, in response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration and the preset transition time. The preset transition time represents the time required for the vehicle to adjust from the real-time acceleration to the target deceleration. Finally, the vehicle is controlled to avoid the target obstacle based on the target deceleration. By co-calculating the lateral avoidance path and the longitudinal braking demand, the feasibility of steering avoidance and braking efficiency are dynamically balanced. This achieves the goal of triggering the automatic emergency steering function when automatic emergency braking is ineffective. Thus, the technical effect of minimizing lateral displacement to achieve the best collision avoidance effect while still achieving collision avoidance is achieved. This solves the technical problem of poor emergency obstacle avoidance effect of assisted driving systems in related technologies.

[0054] Furthermore, controlling the vehicle to avoid a target obstacle based on the target deceleration includes the following steps:

[0055] In response to a target deceleration exceeding a preset deceleration threshold and the existence of steering space, the vehicle is controlled to make an emergency turn to avoid the target obstacle. The preset deceleration threshold is used to represent the upper limit of deceleration within a safe range.

[0056] In this embodiment, the preset deceleration threshold is an upper limit of longitudinal braking capacity set by the driver assistance system based on vehicle dynamics characteristics, active safety function safety levels, occupant comfort boundaries, tire-road adhesion limits, and braking system response capabilities, through analysis of extensive real-vehicle calibration data and simulation conditions. The preset deceleration threshold is not a fixed constant, but rather an adaptive threshold dynamically adjusted based on vehicle speed, road surface adhesion coefficient (e.g., dry / wet / ice / snow), load distribution, tire temperature, and brake fade status. The preset deceleration threshold represents the maximum longitudinal deceleration that the driver assistance system can safely and stably apply without causing unexpected risks such as severe vehicle pitch, tire sideslip, excessive intervention by the electronic stability control system, severe occupant forward movement, or steering system overload.

[0057] Steering avoidance space refers to the physically available space within which a vehicle can safely avoid a target obstacle through lateral movement while in its current driving state. It is not simply a reduction of lane width by half, but rather a three-dimensional safe drivable area constructed by integrating multi-dimensional dynamic constraints. The assessment of steering avoidance space relies on high-precision lane line detection, prediction of the position and dynamics of vehicles in adjacent lanes, boundary identification of static obstacles such as curbs / guardrails / trees, vehicle lateral dynamic response models (e.g., roll angle, yaw inertia), tire side slip characteristics, and the response capability of the steering actuators. The driver assistance system must confirm that there are no other obstacles on the target path, no risk of road boundary encroachment, no vehicles in adjacent lanes approaching dangerously, and that the vehicle's steering response can complete the avoidance trajectory planning and execution within a preset time. The assessment of steering avoidance space also includes "path feasibility verification" and "risk increment assessment," namely, whether the trajectory curvature of the path after steering is within the vehicle's limits, whether the lateral displacement exceeds the safety margin, and whether it will cause a secondary collision or trigger emergency braking by following vehicles. The system will only allow the AES function to be activated and the intelligent decision-making closed loop to be realized when the turning and avoidance space is confirmed to be "available, safe and feasible".

[0058] Responding to a target deceleration exceeding a preset deceleration threshold and the existence of steering space, the system controls the vehicle to perform an emergency steering maneuver to avoid the target obstacle. This can be understood as follows: when the driver assistance system calculates that the longitudinal deceleration required to avoid a collision exceeds the preset deceleration threshold, it indicates that the braking system alone is insufficient to avoid a collision within the physical safety boundaries. At this point, forcibly applying greater braking force may not only trigger the extreme intervention of the anti-lock braking system, leading to tire slippage, vehicle instability, and severe forward lurching of occupants, but may even ultimately result in an inability to avoid a collision due to insufficient braking distance. The feasibility verification of the assisted driving system's subsequent lateral avoidance path was conducted as follows: First, through high-precision lane line recognition and dynamic environment modeling, the system determined in real time whether adjacent lanes were empty, whether there were other moving obstacles, and whether the curb or guardrail was within a safe avoidance boundary. Second, combined with the vehicle dynamics model, the system calculated whether the vehicle could complete a smooth avoidance trajectory without secondary collision risk within a limited lateral space under constraints such as current vehicle speed, steering response delay, tire lateral stiffness, and yaw inertia. Simultaneously, the assisted driving system also assessed the potential impact on traffic behind during the steering process to avoid rear-end collisions caused by sudden lane changes. When it was confirmed that the steering avoidance space met strict constraints in spatial, temporal, and safety dimensions, the system activated the AES function, outputting precise steering angle and torque commands to the steering system. This improved the collision avoidance success rate and system robustness in high-speed complex scenarios, building a safe and reliable redundant collision avoidance capability for advanced intelligent driving.

[0059] It can be seen that by introducing a preset deceleration threshold as the safety boundary of braking capacity, the risks of vehicle instability, tire slippage, or passenger discomfort caused by excessive braking are effectively avoided. At the same time, combined with the multi-dimensional feasibility verification of steering and avoidance space, it is ensured that emergency steering is triggered only under safe and controllable conditions, realizing intelligent coordination between braking and steering, which significantly improves the collision avoidance success rate in complex high-speed scenarios (such as sudden braking of the vehicle in front and pedestrians suddenly appearing), and avoids the technical problem of traditional AEB failing due to insufficient longitudinal space.

[0060] Furthermore, controlling the vehicle to avoid a target obstacle based on the target deceleration includes the following steps:

[0061] In response to a target deceleration being less than or equal to a preset deceleration threshold, the vehicle is controlled to perform emergency braking to avoid the target obstacle.

[0062] In this embodiment, controlling the vehicle to perform emergency braking in response to a target deceleration less than or equal to a preset deceleration threshold to avoid a target obstacle can be understood as the system prioritizing emergency braking when the deceleration required for obstacle avoidance calculated by the driver assistance system is less than or equal to the preset deceleration threshold. Braking does not require changing the vehicle's trajectory, does not cause lane departure risk, does not disturb adjacent traffic flow, and has low sensor dependence, more direct system response, and more reliable execution. The driver assistance system dynamically calculates and requests the maximum but safest deceleration output based on real-time vehicle speed, target distance, road surface adhesion coefficient, and brake thermal state, simultaneously activating cooperative control modules such as the anti-lock braking system to ensure that braking force is precisely applied to each wheel without locking up or skidding, achieving the shortest braking distance. This enables collision avoidance with minimal disturbance and maximum controllability, minimizing the probability of accidents and the severity of their consequences.

[0063] It can be seen that controlling the vehicle to perform emergency braking when the target deceleration is less than or equal to the preset deceleration threshold can fully utilize the physical limits of the vehicle's braking system, achieving the shortest braking distance while ensuring safety and controllability. This is the most direct, stable, and predictable strategy for collision avoidance. This mechanism avoids unnecessary lateral intervention, reduces secondary risks caused by lane changes, and improves the certainty of system response and occupant comfort. Simultaneously, through coordinated chassis control including the anti-lock braking system, it ensures efficient distribution of braking force without locking or skidding, maximizing tire adhesion utilization. This step prioritizes longitudinal control, which not only aligns with the instinctive reactions of human drivers in emergency situations but also provides decision redundancy for subsequent steering and avoidance, significantly improving the system's collision avoidance success rate and reliability in typical scenarios such as high-speed rear-end collisions and sudden pedestrian appearances.

[0064] Furthermore, the vehicle control method also includes the following steps:

[0065] When controlling the vehicle to perform emergency braking, in response to the risk of collision between the vehicle and the target obstacle, a steering collision avoidance point is determined, wherein the steering collision avoidance point is used to indicate the starting position of the steering to safely avoid the target obstacle;

[0066] At the collision avoidance point, steer the vehicle in an emergency to avoid the target obstacle.

[0067] In this embodiment of the application, the steering collision avoidance point refers to the optimal critical position for the start of the lateral avoidance action determined by the driver assistance system during emergency braking, based on multi-dimensional comprehensive calculations including real-time dynamic environmental perception, vehicle kinematics model, target obstacle trajectory prediction, lane environmental constraints, and chassis execution capability.

[0068] For example, when a vehicle decelerates longitudinally to a specific point in time and reaches that point in the longitudinal direction, if relying solely on braking is no longer sufficient to ensure safe avoidance of the target obstacle (such as a vehicle braking suddenly in front or a pedestrian suddenly appearing out of the way), the driver assistance system will initiate an emergency steering command at this precise spatial node. This will guide the vehicle safely to an adjacent lane or a safe avoidance path within the same lane with minimal lateral displacement and the smoothest trajectory change, thus achieving complete collision risk avoidance in both time and space. The steering collision avoidance point is not a fixed preset value, but a real-time decision threshold dynamically recalculated in milliseconds based on parameters such as vehicle speed, target relative speed and acceleration, road curvature, lane availability, distribution of nearby obstacles, tire slip characteristics, steering system response delay, remaining braking capacity, and occupant safety boundaries. This ensures that steering intervention neither prematurely causes unnecessary lane departure and surrounding traffic interference, nor prematurely leads to insufficient avoidance space. It is the key decision boundary for the system to shift from "passive deceleration" to "active avoidance."

[0069] When controlling the vehicle to perform emergency braking, in response to the risk of collision between the vehicle and the target obstacle, determining the steering collision avoidance point can be understood as follows: when controlling the vehicle to perform emergency braking, if there is a risk of collision between the vehicle and the target obstacle, the driver assistance system does not trigger steering at a preset distance or fixed time point. Instead, it continuously integrates multi-source perception data from millimeter-wave radar, cameras, ultrasonic sensors, and high-precision maps to reconstruct the trajectory of the target obstacle in real time, predict its spatial position in the next few seconds, and combine it with the vehicle's current speed, longitudinal deceleration decay curve, lane line geometry, adjacent lane passability, curbs, guardrails, and other environmental constraints. Through high-order kinematics and dynamics models, the steering collision avoidance point is calculated to ensure the safety and reliability of the steering action.

[0070] Controlling the vehicle to perform an emergency steering maneuver at the collision avoidance point to avoid a target obstacle can be understood as follows: upon reaching the collision avoidance point, the driver assistance system outputs an emergency steering command to the electric power steering system, while simultaneously coordinating with the chassis domain controller to apply differential braking force or yaw moment. This assists in quickly establishing body roll and yaw responses, achieving smooth, rapid, and controllable lateral displacement. This steering action is based on a pre-planned optimal avoidance trajectory using closed-loop control, ensuring that the wheel trajectory always stays within the lane boundaries, does not cross the lines, and does not run off the road, thus achieving obstacle avoidance and ensuring that the driver assistance system maintains high reliability and functional safety even under extreme conditions.

[0071] As can be seen, dynamically determining the steering collision avoidance point during emergency braking and triggering the emergency steering function at that point significantly improves the collision avoidance success rate and safety of the assisted driving system in high-risk scenarios. The above steps calculate the steering collision avoidance point by real-time fusion of vehicle dynamics, obstacle trajectory, lane environment, and actuator response characteristics, avoiding false triggering and traffic interference caused by premature steering, and eliminating insufficient avoidance space due to late steering. The dynamic calculation of the steering collision avoidance point ensures that every lateral intervention has sufficient physical feasibility, path safety, and environmental compatibility, effectively avoiding the risks of crossing lane lines, running off the curb, or intruding into adjacent lanes. Simultaneously, the assisted driving system synchronously coordinates steering, braking, and yaw control at the moment of steering initiation, achieving smooth, fast, and low-disturbance avoidance maneuvers, significantly improving passenger comfort and system predictability. This overcomes the limitations of traditional AEB systems that rely solely on longitudinal braking, providing crucial redundancy for extreme scenarios such as high-speed rear-end collisions and sudden pedestrian appearances, significantly enhancing the robustness, reliability, and functional safety level of the intelligent driving system.

[0072] For example, the AES function can plan a reasonable path and achieve collision avoidance when lane markings are clear. When steering to avoid a collision, the AES function minimizes lateral displacement to achieve optimal collision avoidance while still achieving the desired effect. To avoid unexpected accidents caused by function activation, the AES function is not triggered when there is a curb / fence on the target path. When there is an obstacle on the target path, the risk to the target lane is calculated based on the obstacle distance, relative speed, and relative acceleration. If there is a risk to the vehicle, the AES function is not triggered, and only the AEB function is maintained.

[0073] Figure 2 This is a schematic diagram illustrating the triggering of an emergency turn in a high-speed scenario according to an embodiment of this application, such as... Figure 2 As shown, the TTC threshold of the AEB function is greater than the TTC threshold of the AES function (i.e., the AES function is triggered later): This is generally in high-speed conditions. If the AEB function cannot stop the vehicle, the AES function will be triggered. Alternatively, if the AEB function fails to stop the vehicle, it will be interrupted by the AES function, which will then trigger the AES function to achieve the effect of steering to avoid collision.

[0074] Figure 3 It is a schematic diagram of triggering emergency steering in a low-speed scenario according to an embodiment of the present application. As Figure 3 shown, in a low-speed working condition, the TTC threshold of the AEB function < the TTC threshold of the AES function (that is, the AEB function is triggered later), and the AES function is inhibited, and only the AEB function is triggered, which can reduce the false trigger rate of the AES function.

[0075] Furthermore, determining the target deceleration of the vehicle according to the real-time acceleration and the preset transition duration includes the following steps:

[0076] Determining a plurality of acceleration change rates according to the real-time acceleration and the preset transition duration;

[0077] Determining a plurality of decelerations according to the plurality of acceleration change rates and the preset transition duration;

[0078] Determining the deceleration whose absolute value meets the first preset standard among the plurality of decelerations as the target deceleration.

[0079] In an embodiment of the present application, the plurality of acceleration change rates (that is, jerk values, with the unit of m / s³) are a set of discretized and gradient longitudinal acceleration change rate sequences preset by the assisted driving system to achieve a smooth, comfortable and controllable emergency braking transition process, based on the vehicle dynamics model, tire limit characteristics, occupant body feeling threshold and braking system response ability. These acceleration change rates can range from 0.5 m / s³ to 6.0 m / s³, increasing in steps of 0.5 or 1.0 m / s³, covering the complete transition interval from comfortable braking to extreme emergency braking. Each jerk value represents the change rate of the longitudinal acceleration per unit time, and is used to describe the dynamic process of the braking system changing from the current deceleration state to the target deceleration, ensuring that the braking force does not mutate, the body posture does not pitch violently, and the occupants do not have a strong sense of forward rush, while meeting the requirements of "predictability" and "controllability" in the functional safety standard.

[0080] Multiple decelerations can be calculated one by one through integral operations based on the aforementioned multiple acceleration change rates and a unified preset transition duration (i.e., the maximum time window allowed by the system to complete the change from the current deceleration to the target deceleration, typically set to 0.2 to 0.8 seconds, adaptively adjusted according to vehicle speed and scenario risk level). For example, if the current deceleration is 1.0 m / s², and a jerk of 3.0 m / s³ is applied with a transition duration of 0.5 seconds, then the final deceleration increment will be 1.5 m / s², and the final target deceleration will be 2.5 m / s². By traversing all preset jerk values ​​and fixed transition durations, a set of corresponding target decelerations is generated, forming a dynamically selectable "deceleration candidate set" covering the entire feasible range from slight deceleration to extreme braking (e.g., above 8 m / s²).

[0081] The first preset criterion is used to select the target deceleration best suited to the current emergency collision avoidance scenario from the multiple decelerations mentioned above. The first preset criterion is a set of logical judgments that integrate multiple physical and safety constraints. Specifically, the absolute value of the target deceleration should be greater than the "minimum critical collision avoidance deceleration" calculated by the system (i.e., the theoretical minimum value required to avoid a collision by braking alone). The absolute value of the target deceleration must not exceed the maximum usable deceleration allowed by the road surface adhesion coefficient (to prevent tire lock-up or loss of control). The absolute value of the target deceleration must be less than or equal to the maximum output value allowed by the current thermal state and hydraulic capacity of the vehicle's braking system (to prevent brake fade or failure). The absolute value of the target deceleration must meet the occupant comfort boundary. Furthermore, it is necessary to consider the current vehicle speed and target distance to ensure that the selected deceleration can achieve zero collision within the remaining space. Finally, the system selects the minimum absolute value deceleration that meets all constraints, or selects the maximum permissible value in high-risk scenarios, thereby improving the adaptability and robustness of the obstacle avoidance function of the assisted driving system.

[0082] Determining multiple acceleration change rates based on real-time acceleration and preset transition duration can be understood as follows: the driver assistance system does not apply braking force at a fixed rate, but rather constructs multiple possible "acceleration change paths" based on the current longitudinal motion state of the vehicle (i.e., the instantaneous deceleration value collected in real time) and a preset, adaptively adjustable transition time window. The driver assistance system dynamically generates multiple preset jerk (acceleration change rate) candidate values ​​based on current vehicle speed, target distance, road surface adhesion conditions, brake thermal state, and occupant comfort thresholds. For example, multiple acceleration change rates are determined in non-uniform steps from 0.8 m / s³ to 7.0 m / s³, with each acceleration change rate representing a different braking mode. For instance, a low jerk value corresponds to gentle, gradual braking intervention to avoid passenger discomfort from forward lurching, while a high jerk value corresponds to a rapid, decisive emergency response, suitable for extreme scenarios such as high-speed sudden vehicle appearances. These multiple acceleration change rate jerk values ​​are not arbitrarily set but are calibrated based on extensive real-vehicle test data to ensure that functional safety requirements are met and the braking mode expected by the driver is matched under different operating conditions.

[0083] Determining multiple decelerations based on multiple acceleration change rates and preset transition durations can be understood as follows: the driver assistance system takes each preset jerk value as input, integrates it over a fixed time window, and calculates the corresponding longitudinal deceleration value based on fundamental kinematic formulas. For example, when jerk = 2.0 m / s³ and transition duration = 0.5 seconds, the system calculates a deceleration increment of 1.0 m / s². If the current deceleration is 1.5 m / s², the final deceleration will be 2.5 m / s²; if jerk = 6.0 m / s³, a final deceleration of 4.5 m / s² can be achieved under the same duration. By determining multiple decelerations, sufficient decision-making space is provided for subsequent intelligent arbitration.

[0084] Determining the target deceleration as the deceleration whose absolute value meets a first preset standard among multiple decelerations can be understood as the driver assistance system selecting a target deceleration suitable for the current scenario from multiple decelerations based on the first preset standard. For example, the absolute value of the target deceleration should be greater than the "minimum critical collision avoidance deceleration" calculated by the system in real time; otherwise, it is invalid. The absolute value of the target deceleration should be less than or equal to the maximum usable deceleration allowed by the current road surface adhesion coefficient (to prevent tire lock-up, sideslip, or loss of control). The absolute value of the target deceleration should not exceed the upper limit supported by the current heat capacity, hydraulic pressure, and actuator response capability of the braking system (to avoid braking performance degradation due to overload). Furthermore, the occupant's perceived safety boundary must also be considered (e.g., not exceeding 6.8 m / s² to avoid severe forward lurching). If the system detects obstacles, curbs, or oncoming vehicles in adjacent lanes, a more conservative deceleration value is preferentially selected.

[0085] As can be seen, by dynamically constructing a multi-level acceleration change rate (jerk) sequence and combining it with a preset transition duration, a set of candidate deceleration schemes conforming to physical constraints is generated, achieving refined and adaptive control of braking intensity. This step overcomes the shortcomings of traditional fixed deceleration strategies, enabling the assisted driving system to intelligently match the optimal braking curve under different vehicle speeds, road surface adhesion, and risk levels, balancing collision avoidance performance and occupant comfort. It effectively avoids wasted braking force, tire lock-up, or system overload, improves the reliability and predictability of braking response, and enhances control robustness in emergency braking scenarios.

[0086] Furthermore, the vehicle control method also includes the following steps:

[0087] Acquire multiple obstacles around the vehicle;

[0088] Multiple obstacles are filtered based on their positional relationships with the vehicle to obtain at least one initial obstacle;

[0089] By performing lateral and longitudinal risk checks on at least one initial obstacle, target obstacles that meet the second preset criteria are identified.

[0090] In this embodiment, "multiple obstacles" refers to all potential moving or stationary targets in the vehicle's surrounding environment, which are perceived and output in real time by the vehicle's multi-sensor fusion system (including forward-facing millimeter-wave radar, stereo camera, ultrasonic radar, high-precision map, and positioning unit). This includes vehicles, pedestrians, two-wheeled vehicles, traffic cones, guardrails, curbs, construction signs, and low-speed moving objects in the front, side, and rear lanes. Exemplarily, the information of multiple obstacles includes, but is not limited to, target type, confidence level, three-dimensional position coordinates (longitudinal distance, lateral offset, and height relative to the vehicle), velocity vector, acceleration, heading angle, and predicted trajectory; these are not limited here.

[0091] Positional relationship refers to the relative spatial geometric relationship between the vehicle and each obstacle in a two-dimensional plane coordinate system (vehicle coordinate system or world coordinate system). Specifically, it includes: whether the obstacle is located within the longitudinal projection range of the vehicle's current driving lane (i.e., whether it is "within the current lane"), or whether it is located in the lateral adjacent area of ​​an adjacent drivable lane (i.e., whether it is within the "reversible lane space"); whether the lateral offset of the obstacle relative to the vehicle is less than a certain proportion of the preset lane width (e.g., less than 0.8 times the lane width) to determine whether the obstacle poses a direct threat to the vehicle; whether the longitudinal distance of the obstacle is within the effective perception range (e.g., 50–150 meters); and whether the relative speed of the obstacle is close to or greater than the vehicle's speed, thereby determining whether there is a potential risk of rear-end collision or side collision. No restrictions are imposed here.

[0092] The initial obstacle set is a set of candidate targets retained after a first round of spatial and semantic filtering of the aforementioned obstacles. For example, the filtering criteria include: target type validity (excluding low-confidence or meaningless targets such as animals, birds, and irrelevant shadows); whether the target is located in a pre-defined "risk perception sector" in front of or to the side of the vehicle (typically a ±30° lateral view in front, at a distance of 20–120 meters); whether the target is located on a drivable road surface (using high-precision maps and lane line information to eliminate targets in non-traffic areas such as shoulders, green belts, and ditches); and whether the target's perception confidence is higher than a certain threshold (e.g., ≥0.7) to ensure data reliability. After filtering, obstacles possessing both "actual collision probability" and "can be responded to by vehicle actuators" are retained, forming the initial candidate set for subsequent risk assessment.

[0093] Lateral risk verification assesses the navigability of the potential target lane (e.g., the left or right adjacent lane) when a vehicle performs an emergency steering maneuver to avoid introducing new collision risks. This process combines the current lane markings (clarity, continuity, and identifiability) with the presence of other vehicles, curbs, guardrails, fences, construction zones, oncoming traffic, and other static or dynamic obstacles in adjacent lanes to calculate the "lateral safety margin" of the target steering path. If there are insurmountable physical obstacles (e.g., continuous curbs, barriers, oncoming traffic) in the obstacle avoidance path, or if adjacent lanes are completely occupied by other vehicles at excessively high relative speeds, the lateral avoidance path is deemed infeasible, the risk level is increased, and the obstacle is suppressed. Lateral risk verification is essentially a "steering feasibility verification," ensuring that the system does not cause secondary accidents when avoiding current obstacles.

[0094] Longitudinal risk verification assesses whether the collision time between the target obstacle and the vehicle in the current direction of travel is close to or below the system's safety threshold, and whether the current braking capacity is sufficient to achieve effective deceleration within the remaining distance. This process is based on the vehicle's current speed, the target's relative speed and acceleration, combined with the vehicle's braking performance model (including maximum deceleration capacity, braking response delay, tire-road adhesion coefficient, etc.) to calculate whether a collision can be avoided under continuous maximum braking. If the calculated minimum collision avoidance deceleration exceeds the maximum output capacity of the AEB function (e.g., exceeding 7.0 m / s²), or the TTC is less than the preset AES trigger threshold (e.g., 1.2 seconds), the obstacle is determined to be a high-risk target with "insufficient longitudinal avoidance capability." Longitudinal risk verification serves as the basis for determining "whether steering must be initiated," distinguishing between scenarios where "AEB can resolve the issue" and "AEB is insufficient and steering is required."

[0095] The second preset criterion is used to ultimately determine the target obstacle that can trigger the automatic emergency steering (AES) function from the initial obstacles that have undergone dual lateral and longitudinal risk verification. The second preset criterion requires that the target obstacle not only meet the condition of "collision cannot be avoided by braking" in terms of longitudinal risk, but also meet the condition of "a safe avoidance path exists" in terms of lateral risk. Furthermore, it must meet the following constraints: the target type must be a high-risk entity; the target confidence level must be consistently higher than a threshold (to prevent false positives); the target's relative lateral offset must be within the steering correction range (e.g., within ±1.5 meters); the target's predicted trajectory and the vehicle's trajectory must clearly intersect within the time window; and the target must be the "primary threat target" with the highest longitudinal risk and the most feasible lateral path among all candidate obstacles. In addition, system suppression conditions must be excluded, such as driver manual intervention, lane line ambiguity, system function limitations, and severe weather. Ultimately, the target obstacle that simultaneously meets the criteria of high longitudinal risk, lateral avoidability, high confidence level, and highest priority is determined as the target obstacle, serving as a crucial basis for triggering the AES function, fundamentally avoiding false triggering, premature triggering, or invalid triggering.

[0096] Acquiring multiple obstacles around the vehicle can be understood as building a high-precision, high-timeliness, and robust environmental perception model by fusing multi-source heterogeneous perception data from forward-looking cameras, millimeter-wave radar, ultrasonic sensors, surround-view cameras, and high-precision maps. This model continuously scans and identifies all potential dynamic and static targets within a 150-meter radius around the vehicle, including but not limited to decelerating vehicles in the lane ahead, vehicles traveling in the wrong direction in the oncoming lane, pedestrians crossing the road, non-motorized vehicles on bicycles, traffic cones on both sides of the road, construction barriers, curbs, guardrails, bridge piers, stationary vehicles, slow-moving animals, and temporarily appearing construction machinery. Each target is assigned a complete attribute vector, including target type (classification confidence), three-dimensional coordinates (longitudinal distance, lateral offset, and height relative to the vehicle), velocity vector (longitudinal and lateral components), acceleration, heading angle, predicted trajectory, target size, motion stability, and perception fusion confidence, serving as the raw input to the driver assistance system.

[0097] By filtering multiple obstacles based on their positional relationships with the vehicle, at least one initial obstacle can be understood as the driver assistance system performing the first round of intelligent filtering of all targets based on spatial geometric relationships and driving semantic logic. Specifically, firstly, targets located in non-drivable areas, such as green belts, ditches, under overpasses, behind medians, and non-road areas, are excluded. Secondly, low-threat targets located directly behind or to the side of the vehicle with a negative relative speed (i.e., far from the vehicle) are eliminated. Thirdly, only targets located within a ±35° lateral field of view sector in front of the vehicle, with a longitudinal distance within the functionally effective range (20–120 meters), and a lateral offset within ±1.8 meters of the adjacent lane boundary are retained. At the same time, their perception confidence must be consistently higher than a preset threshold (e.g., ≥0.75), and the target type must be a high-risk entity (vehicles, pedestrians, two-wheeled vehicles), while low-value interference items (e.g., birds, billboard shadows, roadside puddles reflections) are excluded. Finally, a small number (usually 1–5) of initial obstacles with real collision potential and operability are selected to avoid invalid calculations.

[0098] By performing lateral and longitudinal risk checks on at least one initial obstacle, the target obstacle that meets the second preset standard can be understood as follows: First, the longitudinal check determines whether the current braking can avoid a collision (e.g., TTC is too short or the required deceleration exceeds the AEB limit). Then, the lateral check verifies whether the steering path is safe (e.g., no obstacles in adjacent lanes, no curbs, no oncoming traffic). Finally, by comprehensively considering the target type, risk level, path feasibility, confidence persistence, and system suppression conditions, an urgent and safe obstacle is selected as the target to trigger the AES function, thus achieving accurate, reliable, and error-free intelligent collision avoidance decision-making.

[0099] As can be seen, by fusing multi-sensor data to acquire multiple obstacles and conducting initial screening based on spatial relationships, irrelevant targets and low-confidence interference are effectively eliminated, significantly reducing computational redundancy and the risk of misjudgment. Furthermore, longitudinal risk verification assesses whether braking is sufficient to avoid collisions, and lateral risk verification verifies the feasibility of the steering path. Finally, based on a second preset standard, only target obstacles that truly meet the dual conditions of "unavoidable braking and safe steering" are triggered, eliminating unexpected steering caused by misjudging curbs, oncoming traffic, or stationary objects, and significantly reducing the probability of secondary accidents. These steps not only improve the adaptability of the assisted driving obstacle avoidance function in complex urban and highway scenarios but also enhance the credibility and user acceptance of the assisted driving system.

[0100] Furthermore, the vehicle control method also includes the following steps:

[0101] Vehicles can make emergency turns based on planned paths, where the planned path is used to avoid target obstacles;

[0102] In response to the presence of a pre-set obstacle on the planned path, the vehicle exits control and makes an emergency turn.

[0103] In this embodiment, the planned path refers to an optimal avoidance trajectory dynamically generated by the driver assistance system after triggering the automatic emergency steering function, based on the current vehicle state (position, speed, heading, yaw rate), the predicted trajectory of the target obstacle, the geometric information of surrounding lane lines, and high-precision map semantic data. This planned path considers a smooth curve of vehicle dynamics constraints (such as maximum steering angle and lateral acceleration limits), tire adhesion, occupant comfort (such as lateral jerk limits), and road environment safety margins. This ensures that the vehicle maintains a safe distance (e.g., ≥0.8 meters) from the target obstacle while minimizing lateral displacement and time, and simultaneously ensures that the vehicle remains within the drivable road surface area.

[0104] Preset obstacles refer to a set of static / dynamic obstacles that are pre-set and detected in real time on or near the planned path, and that are either insurmountable or high-risk, when the driver assistance system performs emergency steering path planning. Preset obstacles include, but are not limited to: continuous curbs in adjacent lanes, concrete guardrails, central dividers, bridge piers, construction barriers, high slopes, ditches, oncoming vehicles (especially on two-way roads without medians), large vehicles that are stationary or moving slowly, and geofences marked as "No Crossing" or "No-Passage Area" on high-precision maps. By fusing camera recognition, radar ranging, map semantic matching, and lane line continuity assessment, the system detects in real time whether the aforementioned obstacles exist on the planned path and evaluates their minimum lateral distance and relative motion trend. Once it is detected that the closest distance between any preset obstacle and the planned path is less than a preset safety threshold (such as 0.5 meters), or if there is a serious risk that would cause a collision, rollover, or running off the road after steering, the system will immediately determine that the path is not feasible, immediately terminate the AES steering control, and return to the AEB braking mode to ensure that the vehicle does not introduce a more serious secondary accident in order to avoid a risk, thereby realizing a safe and proactive obstacle avoidance logic.

[0105] The AES (Automatic Steering) function controls the vehicle's emergency steering based on a planned path. This planned path, used to avoid a target obstacle, can be understood as follows: after longitudinal and lateral risk assessments confirm that braking alone cannot avoid a collision, and there is feasible steering space, the AES function activates the trajectory generation module. Based on the vehicle's real-time status, the predicted trajectory of the target obstacle, the lane geometry provided by a high-precision map, and the drivable area boundaries of the surrounding environment, it generates an optimal avoidance trajectory. This planned path comprehensively considers various factors such as vehicle dynamics limits, tire grip, and passenger comfort (lateral jerk limitations) to ensure a smooth, controllable steering process without severe yaw. This maintains a safe distance (e.g., ≥0.8 meters) between the vehicle and the target obstacle with minimal lateral displacement and in the shortest time, while ensuring the vehicle remains within the drivable road surface throughout the entire process, without crossing lane lines or veering off the road.

[0106] Responding to the presence of a pre-set obstacle on the planned path, the system's emergency steering maneuver can be understood as follows: While the planned path is generated and executed, the system continuously performs dynamic scanning and risk reassessment of the environment ahead, paying particular attention to whether there are insurmountable or high-risk obstacles—i.e., pre-set obstacles—in the areas traversed by the path. Multi-sensor fusion is used to determine the relative position and closest distance between the planned path and the pre-set obstacles in real time. When any pre-set obstacle is detected on the path, and its minimum safe distance from the vehicle's trajectory is lower than the system's set safety threshold (e.g., ≤0.5 meters), or its presence could lead to serious consequences such as vehicle rollover, running off the road, or collision with oncoming vehicles, the system immediately terminates the current steering control command, cancels the emergency torque request sent to the steering system, and simultaneously maintains or enhances AEB braking output to ensure the vehicle safely decelerates within its original lane, avoiding more serious secondary accidents caused by forced steering, and ensuring that the AES function achieves a truly reliable active safety closed loop under all conditions.

[0107] As can be seen, the vehicle's emergency steering based on the planned path ensures effective collision avoidance with minimal lateral displacement, preventing severe yaw and occupant discomfort. Simultaneously, when a pre-set obstacle exists on the planned path, the driver assistance system switches back to AEB braking to prevent more serious risks such as running off the road, rollover, or secondary collisions caused by forced steering. These steps improve the reliability and safety of the obstacle avoidance function in complex real-world scenarios, fully utilizing the steering collision avoidance function while effectively reducing the probability of erroneous intervention.

[0108] Figure 4 This is a schematic diagram of an optional automatic emergency steering process according to an embodiment of this application, such as... Figure 4As shown, firstly, based on the perceived targets output from the upstream perception module, the system actively acquires target information in the environment, and obtains lane line information and the status information of the assisted driving functions through the perception environment model. Next, 32 targets are obtained from the upstream perception fusion module and initially screened (target validity && (target path || target is next to the vehicle)). After the initial screening of AES targets, the AES module filters out targets that are not functionally valid based on target attributes, and verifies the target type, confidence level, and whether the longitudinal distance of the target is within a reliable range, selecting multiple targets. Then, the pre-processed selected targets undergo lateral and longitudinal risk verification, and the target with the highest risk is selected as the primary functional target.

[0109] Specifically, the risk calculation is as follows: Determine the minimum deceleration required for braking to avoid a collision. Avoid a collision by braking, meaning the minimum distance to the target must be greater than zero at any predicted time. Vehicle braking is divided into two phases: a deceleration phase where jerk is constant and a deceleration phase where jerk is zero. Figure 5 This is a schematic diagram of an optional vehicle acceleration curve according to an embodiment of this application, such as... Figure 5 As shown, the displacement of the vehicle over time can be obtained by integration as follows:

[0110]

[0111] x This is the closest longitudinal distance to the obstacle at the current prediction time, which is known. x and t Then the critical condition for:

[0112]

[0113] For each predicted step length for the obstacle, the critical value of the required acceleration is calculated. Based on the current vehicle acceleration value and the transition time calculated through calibration, the final deceleration of the vehicle is calculated. The entire predicted step length is traversed, and the deceleration with the largest absolute value is taken as the final deceleration value to be requested.

[0114] If the calculated collision avoidance deceleration is greater than the maximum deceleration required for AEB to stop the vehicle or for collision avoidance, and there is sufficient space to avoid a collision, then the arbitration triggers the AES function to achieve collision avoidance; otherwise, the AEB function is triggered. During the AEB triggering process, the feasibility of collision avoidance is calculated in real time. If collision avoidance is not possible, the AES function is triggered before the latest turning collision avoidance point of AES to achieve steering collision avoidance.

[0115] The AES system requires status feedback signals from the steering system, including steering wheel angle and rate of change, availability of the steering wheel emergency torque interface, response status of the steering wheel emergency torque interface, and the magnitude of driver torque intervention. These signals are used to determine the driver's driving intentions, vehicle driving status, and steering system status. The steering system responds to emergency torque commands output by the system to ensure the vehicle reaches the AES-planned path to avoid collisions. The AES system uses the braking system for longitudinal braking and may also control the vehicle's direct yaw moment to provide rapid lateral movement. This requires the braking system to support interface responses to ensure vehicle stability and also receives status feedback signals from the braking system. Through braking deceleration, the system reduces speed and lowers the risk of collision. The AES system requires the vehicle's electronic control system to provide a differential drive yaw moment interface (relying on distributed drive) and needs to receive status feedback signals from the powertrain system. Through lateral steering and longitudinal braking, it controls the vehicle to avoid collisions, achieving the best collision avoidance effect.

[0116] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0117] According to an embodiment of this application, a vehicle control device is provided. It should be noted that the device can be used to execute the above-described vehicle control method.

[0118] Figure 6 This is a schematic diagram of a vehicle control device according to an embodiment of this application, such as... Figure 6 As shown, the vehicle control device 600 includes: an acquisition module 601 for acquiring the longitudinal distance between the vehicle and the target obstacle; a determination module 602 for determining the target deceleration of the vehicle based on the real-time acceleration and a preset transition time in response to the longitudinal distance being less than the safe braking distance, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration; and a control module 603 for controlling the vehicle to avoid the target obstacle based on the target deceleration.

[0119] Furthermore, the control module 603 is also used to control the vehicle to make an emergency turn to avoid the target obstacle in response to the target deceleration being greater than a preset deceleration threshold and the existence of steering avoidance space, wherein the preset deceleration threshold is used to represent the upper limit of deceleration within a safe range.

[0120] Furthermore, the control module 603 is also used to control the vehicle to perform emergency braking in response to the target deceleration being less than or equal to a preset deceleration threshold, so as to avoid the target obstacle.

[0121] Furthermore, the control module 603 is also used to determine a steering collision avoidance point in response to a collision risk between the vehicle and the target obstacle when the vehicle is controlled to perform emergency braking, wherein the steering collision avoidance point is used to indicate the starting position of the steering to safely avoid the target obstacle; and to control the vehicle to perform an emergency steering at the steering collision avoidance point to avoid the target obstacle.

[0122] Furthermore, the determining module 602 is also used to determine multiple acceleration change rates based on real-time acceleration and preset transition time; determine multiple decelerations based on multiple acceleration change rates and preset transition time; and determine the deceleration whose absolute value among the multiple decelerations meets a first preset standard as the target deceleration.

[0123] Furthermore, the acquisition module 601 is also used to acquire multiple obstacles around the vehicle; to filter the multiple obstacles according to the positional relationship between the multiple obstacles and the vehicle to obtain at least one initial obstacle; and to determine the target obstacle that meets the second preset standard by performing lateral risk verification and longitudinal risk verification on the at least one initial obstacle.

[0124] Furthermore, the control module 603 is also used to control the vehicle to make an emergency turn based on a planned path, wherein the planned path is used to avoid target obstacles; in response to the presence of a preset obstacle on the planned path, the control module 603 exits control of the vehicle to make an emergency turn.

[0125] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the executable program, wherein the executable program executes the vehicle control method described in any of the above embodiments when running on the processor.

[0126] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0127] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0128] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0129] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0130] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer-readable storage medium, and the computer program is configured to execute the vehicle control method described in any of the above when it is run on a computer or processor.

[0131] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0132] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0133] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0134] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0135] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to perform the vehicle control method of any of the above.

[0136] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0137] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0138] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0139] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0140] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the vehicle control method of various embodiments of this application, including:

[0141] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0142] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0143] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0144] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the vehicle control method in various embodiments of this application, including:

[0145] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0146] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0147] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0148] According to another aspect of the embodiments of this application, a computer program is also provided. When executed by a processor, the computer program implements the vehicle control method in various embodiments of this application, including:

[0149] Step S10: Obtain the longitudinal distance between the vehicle and the target obstacle;

[0150] Step S12: In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and the preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration;

[0151] Step S14: Control the vehicle to avoid the target obstacle based on the target deceleration.

[0152] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0153] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0154] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0155] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0156] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0157] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A vehicle control method, characterized in that, include: Obtain the longitudinal distance between the vehicle and the target obstacle; In response to the longitudinal distance being less than the safe braking distance, the target deceleration of the vehicle is determined based on the real-time acceleration of the vehicle and a preset transition time, wherein the preset transition time is used to represent the time required for the vehicle to adjust from the real-time acceleration to the target deceleration; The vehicle is controlled to avoid the target obstacle based on the target deceleration.

2. The method according to claim 1, characterized in that, The method of controlling the vehicle to avoid the target obstacle based on the target deceleration includes: In response to the target deceleration being greater than a preset deceleration threshold and the existence of steering avoidance space, the vehicle is controlled to make an emergency turn to avoid the target obstacle, wherein the preset deceleration threshold is used to represent the upper limit of deceleration within a safe range.

3. The method according to claim 1, characterized in that, The method of controlling the vehicle to avoid the target obstacle based on the target deceleration includes: In response to the target deceleration being less than or equal to a preset deceleration threshold, the vehicle is controlled to perform emergency braking to avoid the target obstacle.

4. The method according to claim 3, characterized in that, The method further includes: When controlling the vehicle to perform emergency braking, in response to the risk of collision between the vehicle and the target obstacle, a steering collision avoidance point is determined, wherein the steering collision avoidance point is used to indicate the starting position of the steering to safely avoid the target obstacle; At the steering collision avoidance point, the vehicle is controlled to make an emergency turn to avoid the target obstacle.

5. The method according to claim 1, characterized in that, Determining the target deceleration of the vehicle based on its real-time acceleration and a preset transition time includes: Multiple acceleration change rates are determined based on the real-time acceleration and the preset transition duration; Multiple decelerations are determined based on the multiple rates of change of acceleration and the preset transition duration; The deceleration whose absolute value among the plurality of decelerations meets the first preset standard is determined as the target deceleration.

6. The method according to claim 1, characterized in that, The method further includes: Acquire multiple obstacles around the vehicle; Based on the positional relationship between the multiple obstacles and the vehicle, the multiple obstacles are filtered to obtain at least one initial obstacle; By performing lateral and longitudinal risk checks on the at least one initial obstacle, the target obstacle that meets the second preset standard is determined.

7. The method according to claim 2 or 4, characterized in that, The method further includes: The vehicle is controlled to make an emergency turn based on a planned path, wherein the planned path is used to avoid the target obstacle; In response to the presence of a pre-set obstacle on the planned path, the vehicle is deactivated and made to make an emergency turn.

8. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the executable program, wherein the executable program, when run on the processor, performs the vehicle control method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the vehicle control method according to any one of claims 1 to 7 when run on a computer or processor.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the vehicle control method as described in any one of claims 1 to 7.

Images