A method and system for highway collaborative decision-making safe driving
By integrating vehicle-side and roadside perception information on highways, and establishing potential fields for speed, lanes, road boundaries, and physical obstacles, the safety and traffic efficiency issues of autonomous vehicles avoiding emergency obstacles are resolved, resulting in safer and more stable driving performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- DONGFENG MOTOR GRP
- Filing Date
- 2023-05-17
- Publication Date
- 2026-06-26
AI Technical Summary
In highway environments with well-developed vehicle-road cooperative systems, autonomous vehicles avoiding emergency obstacles can easily lead to reduced safety, comfort, and road traffic efficiency for other vehicles on the same road segment.
By acquiring vehicle-side and roadside perception information and fusing it, a velocity potential field, a lane potential field, a road boundary potential field, and a physical obstacle potential field are established. The total potential field is then superimposed to determine the vehicle's driving direction and lateral target acceleration, thereby enabling the vehicle to safely avoid obstacles and drive stably.
It improves the safety of highway driving, reduces traffic accidents, and allows for more stable driving speeds on highways. It also avoids fuel waste and increased exhaust emissions caused by vehicles alternating between fast and slow driving, thereby increasing traffic flow and alleviating traffic congestion.
Smart Images
Figure CN117116072B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of autonomous driving technology, specifically to a method and system for safe driving via highway collaborative decision-making. Background Technology
[0002] Currently, the application and verification scenarios of autonomous vehicles are becoming increasingly widespread. L2 and L3 autonomous vehicles have begun large-scale mass production, while L4 has started specific scenario testing and trial operation verification. However, due to the inherent limitations of sensors such as cameras and radar, the perception capabilities of assisted driving / autonomous driving vehicles are limited. For example, at severely obstructed intersections, they cannot identify whether vehicles are merging from other directions; at traffic light intersections, they cannot recognize traffic light phase information; and collision test accidents occur frequently. Therefore, it is increasingly important to overcome the limitations of single-vehicle perception by combining it with the complex scenarios unique to China, and to improve the safety of assisted / autonomous driving vehicles.
[0003] Intelligent connected sensing and collaborative systems are a key research area in intelligent transportation systems. They primarily involve sharing traffic environment information perceived by vehicles and between vehicles and the roadside, and fusing this data. Compared to single-vehicle intelligent sensing, they offer a wider sensing range and more precise sensing information.
[0004] Existing technology 1 discloses a vehicle collaborative decision-making method, including the following steps: collecting current road condition information and determining the required computing level corresponding to the road condition information; obtaining the available computing power of the master vehicle and determining the master vehicle computing level corresponding to the available computing power; determining whether the master vehicle computing level is greater than or equal to the required computing level; if the master vehicle computing level is less than the required computing level, obtaining the feedback computing power of the peripheral devices corresponding to the master vehicle; and determining the decision result of the master vehicle based on the collaborative calculation between the available computing power and the feedback computing power. The deficiency of existing technology 1 is that it only provides the working method of the collaborative decision-making system, without providing the calculation method for the vehicle's driving state during the collaborative decision-making process; and the given problem of the computing power required for collaborative decision-making is not specific to the specific application scenarios and operating conditions of collaborative decision-making.
[0005] Prior art two provides a vehicle collaborative decision-making method, device, electronic device, and computer storage medium. This vehicle collaborative decision-making method, applied to a cloud server, includes: receiving a collaborative decision-making request sent by a roadside device; wherein the collaborative decision-making request is sent by the roadside device after acquiring real-time road information and determining that road congestion has occurred based on the real-time road information; acquiring road scene information attached to the collaborative decision-making request; determining the congestion scene type based on the road scene information; calculating a multi-vehicle decision-making plan scheme using a preset scene collaborative decision-making model based on the congestion scene type; and sending the corresponding decision-making plan scheme to each vehicle, so that each vehicle executes the corresponding driving operation according to its respective decision-making plan scheme. The deficiency of prior art two is that while it provides the working method of the collaborative decision-making system, it does not provide the calculation method for the vehicle's driving state during the collaborative decision-making process; it mainly provides the system architecture of the collaborative decision-making system, and the working relationships between the subsystems under this architecture, but does not provide the changes in vehicle operating state and driving state after collaborative decision-making in various scenarios. Summary of the Invention
[0006] The purpose of this invention is to provide a method and system for safe driving through highways with cooperative decision-making, in order to solve the problem that when autonomous vehicles avoid emergency obstacle sections in highway environments with good vehicle-road cooperation conditions, it can easily lead to a decrease in the driving safety and comfort of other vehicles on the road and a reduction in road traffic efficiency.
[0007] To address the aforementioned technical problems, this invention provides a technical solution: a method for safe driving via highway collaborative decision-making, comprising the following steps:
[0008] S1. Acquire vehicle-side perception information and roadside perception information and perform information fusion; the vehicle-side perception information is acquired through on-board sensors, and the roadside perception information is acquired through roadside sensors.
[0009] S2. Establish a plane rectangular coordinate system on the highway surface, and establish a velocity potential field, lane potential field, road boundary potential field, and physical obstacle potential field on the road surface based on the fused perception information; among them, the velocity potential field makes the vehicle tend to travel at the set desired speed, the lane potential field makes the vehicle tend to stay in the current lane, the road boundary potential field makes the vehicle tend to travel away from the road boundary, and the physical obstacle potential field makes the vehicle tend to travel away from the physical obstacle.
[0010] S3. The vehicle is affected by the total potential field formed by the superposition of the velocity potential field, the lane potential field, the road boundary potential field, and the physical obstacle potential field. Based on the potential field intensity gradient of the total potential field along the driving direction and the lateral direction at the vehicle position, the target acceleration of the vehicle in the driving direction and the lateral direction is obtained.
[0011] S4. The vehicle travels according to the target acceleration in the direction of travel and the lateral direction.
[0012] According to the above scheme, the plane rectangular coordinate system takes the direction of road travel as the positive x-axis and the direction perpendicular to the x-axis and pointing inwards from the road as the positive y-axis.
[0013] According to the above scheme, the velocity potential field Represented as,
[0014]
[0015] In the above formula, The set velocity potential field coefficient, The component of the vehicle's velocity on the x-axis. The desired speed is set;
[0016] Lane momentum field Represented as,
[0017]
[0018] In the above formula, The set lane potential field coefficient, Lane width;
[0019] Road boundary potential field Represented as,
[0020]
[0021] In the above formula, The potential field coefficients of the road boundary are set, and n is the total number of lanes in the current road travel direction;
[0022] Physical obstacle potential field Represented as,
[0023]
[0024] In the above formula, The coefficient of the potential field of the solid obstacle. The distance between the vehicle and the physical obstacle. To set parameters.
[0025] According to the above scheme, a longitudinal control model for the vehicle is established based on the gradient of the total potential field in the vehicle's driving direction. Specifically,
[0026] Since neither the lane potential field nor the road boundary potential field has a gradient change in the x-direction, the vehicle longitudinal control model only considers the influence of the velocity potential field and the potential field of the physical obstacle.
[0027] When the vehicle When there is no lane change and no intention to change lanes, it affects vehicles. Potential field of velocity in the x direction Represented as,
[0028]
[0029] In the above formula, Indicates vehicle The corresponding velocity potential field is affected by the vehicle. driving speed Influence; Represented as,
[0030]
[0031] In the above formula, For vehicles In the vehicle The strength of the potential field formed by the physical obstacle at that location. A collection of vehicles within a certain road section. For static obstacles In the vehicle The strength of the potential field formed by the physical barrier at that location. A collection of stationary obstacles within a certain road section; and The possible values are as follows:
[0032]
[0033]
[0034] In the above formula, for x-axis coordinates For vehicles x-axis coordinates For static obstacles x-axis coordinates The set distance value;
[0035] At this moment, the vehicle's directional acceleration for,
[0036]
[0037] In the above formula, This is the scaling factor between the gradient and the acceleration in the direction of travel;
[0038] When a vehicle changes lanes, the vehicle's coordinates are: At this point, the potential field affecting the vehicle's velocity in the x-direction is represented as:
[0039]
[0040] In the above formula, Let the vehicle's y-axis coordinate be The potential field affecting the vehicle's velocity in the x-direction when traveling along the lane centerline. Let the vehicle's y-axis coordinate be The potential field affecting the vehicle's velocity in the x-direction when traveling along the lane centerline. This represents the y-axis coordinate of the lane centerline of the lane the vehicle was in before changing lanes. The y-axis coordinate of the lane centerline where the vehicle is located after changing lanes; The lane change completion rate is specifically represented as follows:
[0041]
[0042] At this moment, the vehicle's directional acceleration for,
[0043] .
[0044] According to the above scheme, a longitudinal control model for the vehicle is established based on the gradient of the total potential field in the lateral direction of the vehicle. Specifically...
[0045] First, determine the lane change motivation; when the driving speed is... At that time, for vehicles that are not in the process of changing lanes When it satisfies At that time, it was assumed that the vehicle was in the lane. Change lanes to the adjacent lane. It can achieve a more ideal driving state; among which For lane The centerline y-axis coordinate, Lane The centerline y-axis coordinate, and They are vehicles In coordinates and coordinates The total potential field strength at that time, The set lane-changing threshold;
[0046] Then, a lane-changing safety assessment is performed; when a vehicle has the intention to change lanes to an adjacent lane, if there are vehicles traveling behind the vehicle changing lanes in the target lane. When the following conditions are met, the lane-changing safety is deemed met, and the lane change is executed.
[0047]
[0048] In the above formula, For vehicles driving speed, This represents the maximum deceleration of vehicles changing lanes. For vehicles x-axis coordinates For vehicle body length, For the speed of vehicles changing lanes, For vehicles Maximum deceleration, The x-axis coordinate of the vehicle changing lanes;
[0049] When a vehicle changes lanes, its lateral acceleration for,
[0050]
[0051] In the above formula The set time length; before the lane-changing vehicle passes the lane line between the target lane and the original lane, the lateral acceleration direction is towards the target lane; after the lane-changing vehicle passes the lane line between the target lane and the original lane, the lateral acceleration direction is towards the original lane.
[0052] According to the above scheme, when an emergency obstacle is detected in the current lane while the vehicle is driving, the vehicle will avoid the obstacle according to the emergency obstacle avoidance model. Specifically,
[0053] The vehicle decelerates under the influence of the physical obstacle potential field generated by the obstacle and continuously assesses the safety of lane changing in adjacent lanes. When there is an adjacent lane that meets the safety requirements for lane changing, the vehicle changes lanes. During the lane change process, a virtual potential field is generated in parallel on the target lane to replace the physical obstacle potential field currently generated by the lane-changing vehicle.
[0054] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement steps S2 to S3 of the highway collaborative decision-making safe driving method described above.
[0055] A computer-readable storage medium having a computer program stored thereon, wherein when executed by a processor, the computer program implements steps S2 to S3 of the highway collaborative decision-making safe driving method described above.
[0056] A car that uses the highway collaborative decision-making safe driving method described above for assisted driving.
[0057] A highway collaborative decision-making safe driving system for implementing the highway collaborative decision-making safe driving method described above includes,
[0058] The cloud subsystem is used to integrate vehicle-side perception information and roadside perception information, and generate vehicle driving decisions based on the integrated perception information.
[0059] The roadside subsystem is used to acquire roadside sensing information and relay data between the cloud subsystem and the vehicle terminal subsystem.
[0060] The vehicle terminal system is used to acquire vehicle-side perception information and control vehicle movement based on vehicle driving decisions.
[0061] The beneficial effects of this invention are as follows: By acquiring and fusing sensory information from vehicles and road sections, a velocity potential field, a lane potential field, a road boundary potential field, and a physical obstacle potential field are established, which are then superimposed to obtain a total potential field. Based on the potential field intensity gradient along the driving direction and laterally at the vehicle's position within the total potential field, the vehicle obtains its target acceleration in the driving direction and laterally, achieving a safer and more efficient highway driving effect. This solution can improve highway driving safety and reduce traffic accidents. Simultaneously, this solution enables vehicles to travel at more stable speeds on highways, avoiding the fuel waste and increased exhaust emissions caused by alternating high and low speeds. Furthermore, using this method on highways can increase traffic flow, alleviate traffic congestion, and reduce travel time. Attached Figure Description
[0062] Figure 1 This is a structural diagram of a highway collaborative decision-making safe driving system according to an embodiment of the present invention;
[0063] Figure 2 This is a schematic diagram of coordinate system establishment according to an embodiment of the present invention;
[0064] Figure 3 This is a schematic diagram of a vehicle lane changing according to an embodiment of the present invention;
[0065] Figure 4 This is a schematic diagram of a vehicle lane-changing path curve according to an embodiment of the present invention;
[0066] Figure 5 This is a schematic diagram of a vehicle emergency obstacle avoidance scenario according to an embodiment of the present invention;
[0067] Figure 6 This is a flowchart of a highway collaborative decision-making safety driving process according to an embodiment of the present invention. Detailed Implementation
[0068] To make the objectives, technical solutions, and advantages of the embodiments of this disclosure clearer, the technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this disclosure. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure without creative effort are within the scope of protection of this disclosure.
[0069] See Figure 1The vehicle-road cooperative system mainly comprises three subsystems: the cloud subsystem, the roadside subsystem, and the vehicle-to-everything (V2X) subsystem. The cloud subsystem, as the central hub of the V2X system, acquires various data sources (roadside device perception data, vehicle operation data, and vehicle-to-everything sensor perception data), performs real-time data fusion processing, and finally makes macro-level decisions and controls vehicle behavior within the system based on the fusion results. The roadside subsystem includes various radar and camera sensing devices, which can perceive vehicle and traffic conditions in real time and also serve as an information bridge between the cloud subsystem and the V2X subsystem. The V2X subsystem serves individual autonomous vehicles, possesses necessary sensing devices, and controls the vehicle based on decisions made by the cloud system.
[0070] Under optimal vehicle-road cooperation conditions on highways, vehicle driving behavior exhibits the following characteristics:
[0071] ① All vehicles are autonomous vehicles and meet the following requirements. Figure 1 System architecture;
[0072] ② All vehicles travel in the same direction and along the highway's forward direction. Reversing or going in the opposite direction is not allowed. The speed of all vehicles is relatively even and close to the highway's maximum speed limit, but will not exceed the maximum speed limit.
[0073] ③ When a vehicle is driving normally on a highway, it tends to stay along the center line of its lane. Only when a vehicle intends to change lanes will it leave its current lane, cross the lane line, and continue driving along the center line of the target lane.
[0074] ④ When a vehicle is traveling on a highway, the most important factor affecting its speed is the traffic conditions in front of the vehicle, including the driving conditions of the leading vehicle in the same lane and obstacles that suddenly appear in front of it. The driving conditions of the following vehicles and other vehicles in the side lanes have a relatively small impact on the vehicle's driving.
[0075] The first phase establishes an evaluation system for the impact on all vehicles in a highway environment with well-developed vehicle-road cooperative systems. This scheme uses an artificial potential field to describe the factors affecting vehicle driving states and provides the corresponding potential field strength function. Considering the characteristics of vehicle driving on highways, Michael T. Wolf and Joel W. Burdick proposed a set of artificial potential field functions suitable for highways in 2008 to describe the motion of autonomous vehicles on multi-lane highways. The potential field on highways is divided into four types: velocity potential, lane potential, road boundary potential, and physical obstacle potential. This considers not only the influence of other vehicles but also other physical obstacles, such as accident zones and obstacles unexpectedly entering the road, thus reasonably reflecting changes in the potential field. Total potential field strength. It can be represented as:
[0076]
[0077] in, These are the potential field strengths of velocity, lane, road boundary, and physical obstacle, respectively.
[0078] The second stage involves establishing a reasonable coordinate system. Since highways are not perfectly straight, the direction of the highway lane lines is chosen as the coordinate system. The positive axis direction, i.e., the longitudinal direction of vehicle travel, is perpendicular to the lane lines. The positive axis direction represents the lateral direction of vehicle travel. Minor deformations due to highway curvature and road undulations are ignored. The origin of the coordinate system is chosen as the outer edge of the rightmost lane at the starting point of the studied road segment. To simplify calculations, the vehicle's position is represented as a rectangle, with the center of the rectangle as the vehicle's coordinate axis. Location, such as Figure 2 As shown.
[0079] The third stage establishes the velocity potential function. When a vehicle travels on a highway, its direction of travel is always along the road's forward direction, and there exists a desired velocity. (Set by the system or according to user preference) Under normal circumstances, the vehicle speed will tend towards the desired speed. Speed less than the maximum speed limit of the highway Under conditions of no external interference, if the vehicle's actual speed... The vehicle will accelerate to And continue driving at that expected speed; if the actual speed of the vehicle is greater than... The vehicle will slow down to And continue driving at that desired speed.
[0080] When the actual speed differs significantly from the desired speed, the vehicle will approach the desired speed with greater acceleration or deceleration. Conversely, when the difference is smaller, the vehicle will gradually adjust its speed to approach the desired speed. Therefore, a linear relationship can be used to describe the velocity potential:
[0081]
[0082] in, This represents the component of the current vehicle speed in the direction of travel along the road. This is the velocity potential coefficient. To simplify the system's computational complexity, its value can be set to 1, based on the desired velocity. And vehicle comfort acceleration calibration settings.
[0083] Under the influence of velocity potential, the vehicle The relationship between acceleration and current velocity is as follows:
[0084]
[0085] in, For the longitudinal acceleration of the vehicle, For the vehicle's lateral acceleration, The scaling factor between the gradient and acceleration / deceleration is used in this scheme, based on the vehicle's maximum deceleration and... The settings are configured accordingly. Therefore, without considering other factors, the velocity potential has no effect on the vehicle's lateral acceleration. The larger the value, the stronger the vehicle's desire to approach the desired speed, and the shorter the time it takes to reach the desired speed. However, The larger the vehicle, the less comfortable it will be for passengers.
[0086] In the fourth stage, a lane potential function is established, and vehicles tend to travel along the current lane centerline. Therefore, a virtual lane potential function can be set at the lane lines to prevent vehicles from arbitrarily crossing the lane centerline. However, it is also necessary to ensure that vehicles can overcome the lane potential function and switch to alternative lanes when necessary, such as to overcome obstacles ahead or seek a better driving state. For this purpose, a sinusoidal lane potential function is set at the lane lines:
[0087]
[0088] in, The lane line potential coefficient can be set to 1 to simplify the system's calculation complexity. It reflects the strength of the lane line's obstruction of vehicles changing lanes. The larger the lane, the more difficult it is to meet the lane-changing conditions, and the harder it is for vehicles to overcome lane obstacles to change lanes. This refers to the lane width.
[0089] The fifth stage establishes the road boundary potential function. Except for the lane lines which impede the lateral movement of vehicles, vehicles cannot cross the left and right boundaries of the road during normal driving. Therefore, it is necessary to set relatively large road boundary potentials near the lane boundaries, with the potential increasing closer to the road boundary and approaching infinity as the distance from the boundary decreases.
[0090]
[0091] in, This is the road boundary potential coefficient. To simplify system calculations, a value of 1 can be set. It reflects the resistance strength of the road boundary to approaching vehicles. The larger the value, the more likely the vehicle is to drive away from the road boundary; This refers to the number of lanes.
[0092] The sixth stage establishes the actual obstacle potential function. Physical obstacles on highways include other vehicles traveling normally and emergency obstacles in special circumstances, such as accident vehicles or stray animals. The obstacle potential should continuously increase as a vehicle approaches the obstacle. Simultaneously, the obstacle potential increases slowly when the target vehicle is far from the obstacle, but increases rapidly when the target vehicle is close. Therefore, when selecting the obstacle potential field function, it is necessary to ensure that the rate of change of the function increases sharply as the distance difference decreases. Here, the Yukawa Potential field strength function, used to describe the short-range interaction strength between nucleons, can be chosen:
[0093]
[0094] in, This is the entity obstacle potential coefficient; to simplify the system's computational complexity, a value of 100 can be set. The distance between the target vehicle and the physical obstacle can be directly sensed and acquired by the system's perception system. For model parameters, a value of 0.5 can be set to simplify system computational complexity; reaction Follow The magnitude of the rate of change. Choosing this function as the entity barrier potential function meets our requirements, namely, the closer the distance, the higher the field strength and the faster it increases.
[0095] The seventh stage establishes a single-vehicle lateral and longitudinal control model. Cloud-based longitudinal control of the vehicle refers to its longitudinal acceleration and deceleration. Only two types of potential fields act on the vehicle and exhibit gradient changes in intensity along the road direction: velocity potential and physical obstacle potential. Lane potential and road boundary potential... There is no gradient change in any direction, that is:
[0096]
[0097]
[0098] When the vehicle speed greater than expected speed At that time, the direction of the velocity potential is along In the negative direction of the axis,
[0099] When the vehicle speed Less than the expected speed At that time, the direction of the velocity potential is along Positive axis direction
[0100] In a well-developed vehicle-road cooperative environment, the cloud-based decision-making system possesses global perception information. Therefore, it can be reasonably assumed that a vehicle traveling normally in a certain lane does not affect the operation of vehicles in other lanes. Here, "vehicles traveling normally" refers to vehicles traveling in their own lane, not performing lane-changing operations, and showing no intention to change lanes. When the longitudinal positional difference between the target vehicle and the physical obstacle is less than... Only then is it considered that a physical obstacle affects the target vehicle. Therefore, for a vehicle in the lane that is not in the process of changing lanes... Potential field at the location of The strength is:
[0101]
[0102] in, , For vehicles In the vehicle The strength of the physical obstacle potential caused by its location. This refers to the collection of all vehicles on this section of road. For static obstacles In the vehicle The strength of the physical obstacle potential caused by its location. This is the set of all stationary obstacles on this road segment. The possible values are as follows:
[0103]
[0104]
[0105] For a vehicle changing lanes, the traffic conditions in both the original and target lanes affect its longitudinal acceleration. During the lane change, the influence of vehicles and obstacles in both lanes changes gradually: initially, the target vehicle is more affected by the original lane; as the lane change nears completion, it is primarily affected by the target lane. When a vehicle is between two lanes, the potential field strength at its current location can be calculated based on the lane change completion rate. The lane change completion rate is defined as the ratio of the lateral distance traveled by the vehicle from the start of the lane change to the current moment to the lateral distance required to complete the lane change. Note that vehicles tend to travel along the lane center; therefore, the lateral distance required to complete the lane change is the lane width. Lane Change Completion Rate The calculation formula is:
[0106]
[0107] in, This represents the lateral coordinates of the vehicle during the lane change process. The coordinates of the original lane centerline. Lane width. Lane change completion rate. The value ranges from 0 to 1.
[0108] See Figure 3 The vehicle is in the lane Change lanes During the process, we first assume that the vehicle is located in both the original lane and the target lane, with the horizontal axis representing the vehicle's current coordinates. The vertical axis represents the coordinates of the original lane centerline. Coordinates of the center line of the target lane The potential field strength of the vehicle at the two virtual positions is calculated separately, and then the sum of the current actual velocity potential and the physical obstacle potential strength is calculated proportionally based on the lane change completion rate.
[0109]
[0110] in, To improve the lane change completion rate, Assuming the vehicle is in the original lane When on the center line The total potential field strength at the coordinates, Assuming the vehicle is in the target lane When on the center line The total potential field strength at the coordinates.
[0111] Under these conditions, the vehicle's longitudinal acceleration and deceleration are determined by... along The direction is determined by the gradient. That is:
[0112]
[0113] in, This is the scaling factor between the gradient and the acceleration / deceleration.
[0114] The vehicle lateral control problem is essentially a lane-changing decision-making problem, addressing whether to change lanes, when to change lanes, and in which direction to change lanes. This decision-making process considers two factors: lane-changing motivation and lane-changing safety. Lane-changing motivation refers to whether a vehicle should change lanes to achieve a better driving state. The cloud analyzes the current vehicle's lane and the driving status of vehicles in adjacent lanes, as well as obstacle conditions, to determine whether the vehicle should change lanes at the current moment, and whether to change lanes to the left or right. Therefore, the primary consideration in generating lane-changing motivation is whether the target vehicle can more easily achieve higher acceleration to reach its ideal speed or decelerate with less deceleration in the target lane. Lane-changing decisions are only meaningful for vehicles in a normal driving state (not in the process of changing lanes). Therefore, the target vehicle can be considered to be located in the center of its current lane, i.e. .
[0115] when At that time, for vehicles that were not in the lane-changing process A vehicle can be considered to have changed lanes when the following conditions are met. This will allow the vehicle to achieve a more ideal driving condition:
[0116]
[0117] in, The total field strength at the vehicle's current position. Assuming the vehicle is on the center line of the target lane The total field strength at the coordinates, and Note that, due to Since the coordinates are the same and the velocities used to calculate the velocity potential are both the current velocities of the vehicles, the velocity potentials of the vehicles at the current position and the corresponding positions on the target lane are equal.
[0118]
[0119] Regarding road boundary potential, when a vehicle is located at the center line of the outermost or innermost lane...
[0120]
[0121] Regarding lane position, when the vehicle is in the center line,
[0122]
[0123] Therefore, the previous equation can be approximated as:
[0124]
[0125] In other words, by simply comparing the potential field strength of the physical obstacles at the two locations, it can be determined whether the driving state of the target lane is better than that of the current lane. If the potential field strength of the physical obstacles in the side lane is lower than that of the current lane, there is an incentive to switch lanes to the side lane; if the potential field strength of the physical obstacles in the side lane is higher than that of the current lane, it is better to stay in the current lane.
[0126] In reality, the potential field strengths on two lanes can be very similar; even small changes in vehicle speed or relative position can cause the potential field strengths to interchange. Using this as the standard for determining whether to change lanes could lead to excessively frequent lane changes. Therefore, a threshold needs to be added. To ensure that the vehicle does not overreact to changes in the potential field strength:
[0127]
[0128] When the above conditions are met, it is considered that there is a lane-changing motivation. When the vehicle is not in the innermost or outermost lane, it is necessary to consider the vehicle's left and right lanes separately, that is, to compare them separately. and The magnitudes of the potential energy fields of the physical obstacles in the current lane and the physical obstacles in the adjacent lane.
[0129] The eighth stage of lane-change safety analysis, after determining that the vehicle has the motivation to change lanes to at least one side, needs to consider whether the vehicle can safely change lanes. This assumes that the vehicle in the target lane is longitudinally behind the target vehicle and is the closest to it. Its coordinates are The current speed is The target vehicle's coordinates are The current speed is To ensure safety, it is assumed that the current vehicle will immediately change lanes and decelerate at maximum speed during the lane change process. Brake to a stop. If simultaneously at maximum deceleration Braking must be able to prevent a collision with the vehicle in front. That is:
[0130]
[0131] in, This refers to the length of the vehicle.
[0132] If a vehicle has the motivation to change lanes and meets the safety requirements for lane changing, the cloud system can consider it reasonable for the vehicle to change lanes in that direction. If the vehicle is in the innermost or outermost lane, it can only change lanes in one direction. In this case, it is only necessary to determine the motivation and safety of lane changing on one side of the lane. If the conditions are met, a lane changing instruction can be issued to the vehicle. If the vehicle is in the middle lane and both the left and right lanes meet the lane changing conditions, the potential field strengths of the two lanes are compared, and the lane with the lower field strength is selected as the target lane, and a lane changing instruction is issued to the vehicle.
[0133] During a lane change, the relationship can be approximated as a time-lateral displacement relationship, such as... Figure 4 As shown, the vehicle From time From the lane The vehicle begins to change lanes at the center line. along The velocity in the axial direction is 0, and after a time... (Typically set to 5 seconds) Complete lane change, lateral position reaches the lane. At the center of the vehicle at that moment along The speed in the axial direction is 0. The vehicle then continues along the lane. Traveling along the centerline. The entire lateral motion can be divided into two segments: uniformly accelerated motion and uniformly decelerated motion. That is:
[0134]
[0135] The ninth stage is the analysis of collaborative decision-making scenarios for emergency obstacle avoidance. In emergency obstacle avoidance situations, the vehicle-road cooperative system detects an emergency obstacle at a certain location on the road, such as... Figure 5 As shown, for vehicles behind the obstacle A lane-changing maneuver is required to bypass the obstacle and continue driving. Entering the obstacle After the affected area is reached, longitudinal control requires deceleration, and lateral control requires determining whether lane changing to an alternate lane is possible. If the traffic conditions in the alternate lane consistently fail to meet safety requirements, the vehicle may decelerate to a standstill, waiting for a lane-changing opportunity to appear. For vehicles... Vehicles hoping to change lanes to the side lane In other words, it is necessary to ensure the vehicle During lane change and vehicles Safety is paramount. To ensure traffic efficiency, vehicle safety, and passenger comfort on this road section, additional rules are needed to meet the requirements of multi-vehicle cooperative control.
[0136] After the target vehicle begins its lane-changing maneuver, a virtual potential field is generated at the corresponding location in the target lane to replace the virtual potential field generated at the target vehicle's current location, thus facilitating the movement of vehicles in the target lane. and its following vehicles Make a reasonable response before the target vehicle completely changes into the lane.
[0137] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.
Claims
1. A method for safe driving via highway collaborative decision-making, characterized in that: Includes the following steps, S1. Acquire vehicle-side perception information and roadside perception information and perform information fusion; the vehicle-side perception information is acquired through on-board sensors, and the roadside perception information is acquired through roadside sensors. S2. Establish a plane rectangular coordinate system on the highway surface, and establish a velocity potential field, lane potential field, road boundary potential field, and physical obstacle potential field on the road surface based on the fused perception information; Among them, the velocity potential field tends the vehicle to travel at the set desired speed, the lane potential field tends the vehicle to stay in the current lane, the road boundary potential field tends the vehicle to travel away from the road boundary, and the physical obstacle potential field tends the vehicle to travel away from the physical obstacle. S3. The vehicle is affected by the total potential field formed by the superposition of the velocity potential field, the lane potential field, the road boundary potential field, and the physical obstacle potential field. Based on the potential field intensity gradient of the total potential field along the driving direction and the lateral direction at the vehicle position, the target acceleration of the vehicle in the driving direction and the lateral direction is obtained. S4. The vehicle travels according to the target acceleration in the direction of travel and the lateral direction; Velocity potential field Represented as, In the above formula, The set velocity potential field coefficient, The component of the vehicle's velocity on the x-axis. For the set desired speed, The x-axis coordinate of the vehicle changing lanes; Lane momentum field Represented as, In the above formula, The set lane potential field coefficient, Lane width; Road boundary potential field Represented as, In the above formula, Here, n represents the potential field coefficients at the road boundary, and n is the total number of lanes in the current road travel direction. Let y be the coordinate of the vehicle changing lanes; Physical obstacle potential field Represented as, In the above formula, The coefficient of the potential field of the solid obstacle. The distance between the vehicle and the physical obstacle. To set parameters.
2. The highway collaborative decision-making method for safe driving according to claim 1, characterized in that: The Cartesian coordinate system uses the direction of road travel as the positive x-axis and the direction perpendicular to the x-axis and pointing inwards from the road as the positive y-axis.
3. The highway collaborative decision-making method for safe driving according to claim 1, characterized in that: Based on the gradient of the total potential field in the vehicle's direction of travel, a longitudinal control model for the vehicle is established. Specifically, Since neither the lane potential field nor the road boundary potential field has a gradient change in the x-direction, the vehicle longitudinal control model only considers the influence of the velocity potential field and the potential field of the physical obstacle. When the vehicle When there is no lane change and no intention to change lanes, it affects vehicles. Potential field of velocity in the x direction Represented as, In the above formula, Indicates vehicle The corresponding velocity potential field is affected by the vehicle. driving speed Influence; Represented as, In the above formula, For vehicles In the vehicle The strength of the potential field formed by the physical obstacle at that location. A collection of vehicles within a certain road section. For static obstacles In the vehicle The strength of the potential field formed by the physical obstacle at that location. A collection of stationary obstacles within a certain road section; and The possible values are as follows: In the above formula, for x-axis coordinates For vehicles x-axis coordinates For static obstacles x-axis coordinates The set distance value; At this moment, the vehicle's directional acceleration for, In the above formula, This is the scaling factor between the gradient and the acceleration in the direction of travel; When a vehicle changes lanes, the vehicle's coordinates are: At this point, the potential field affecting the vehicle's velocity in the x-direction is represented as: In the above formula, Let the vehicle's y-axis coordinate be The potential field affecting the vehicle's velocity in the x-direction when traveling along the lane centerline. Let the vehicle's y-axis coordinate be The potential field affecting the vehicle's velocity in the x-direction when traveling along the lane centerline. This represents the y-axis coordinate of the lane centerline of the lane the vehicle was in before changing lanes. The y-axis coordinate of the lane centerline where the vehicle is located after changing lanes; The lane change completion rate is specifically represented as follows: At this moment, the vehicle's directional acceleration for, 。 4. The highway collaborative decision-making method for safe driving according to claim 1, characterized in that: Based on the gradient of the total potential field in the lateral direction of the vehicle, a longitudinal control model for the vehicle is established. Specifically... First, determine the lane change motivation; when the driving speed is... At that time, for vehicles that are not in the process of changing lanes When it satisfies At that time, it was assumed that the vehicle was in the lane. Change lanes to the adjacent lane. It can achieve a more ideal driving state; among which For lane The centerline y-axis coordinate, Lane The centerline y-axis coordinate, and They are vehicles In coordinates and coordinates The total potential field strength at that time, The set lane-changing threshold; Then, a lane-changing safety assessment is performed; when a vehicle has the intention to change lanes to an adjacent lane, if there are vehicles traveling behind the vehicle changing lanes in the target lane. If the following conditions are met, the lane-changing safety is deemed met, and the lane change is executed. In the above formula, For vehicles driving speed, The maximum deceleration of vehicles changing lanes. For vehicles x-axis coordinates For vehicle body length, For the speed of vehicles changing lanes, For vehicles The maximum deceleration; When a vehicle changes lanes, its lateral acceleration for, In the above formula The set time length; Before a vehicle changes lanes, its lateral acceleration is directed toward the target lane. After the vehicle changes lanes, its lateral acceleration is directed toward the original lane.
5. The highway collaborative decision-making safe driving method according to claim 4, characterized in that: When a vehicle detects an emergency obstacle in its current lane, it avoids the obstacle according to an emergency obstacle avoidance model. Specifically, The vehicle decelerates under the influence of the physical obstacle potential field generated by the obstacle and continuously assesses the safety of lane changing in adjacent lanes. When there is an adjacent lane that meets the safety requirements for lane changing, the vehicle changes lanes. During the lane change process, a virtual potential field is generated in parallel on the target lane to replace the physical obstacle potential field currently generated by the lane-changing vehicle.
6. A computer device, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement steps S2 to S3 of the highway collaborative decision-making safe driving method according to any one of claims 1-5.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements steps S2 to S3 in the highway collaborative decision-making safe driving method according to any one of claims 1-5.
8. A car, characterized in that: It employs the highway collaborative decision-making safe driving method described in any one of claims 1-5 for assisted driving.
9. A highway collaborative decision-making safe driving system for implementing the highway collaborative decision-making safe driving method according to any one of claims 1-5, characterized in that: include, The cloud subsystem is used to integrate vehicle-side perception information and roadside perception information, and generate vehicle driving decisions based on the integrated perception information. The roadside subsystem is used to acquire roadside sensing information and relay data between the cloud subsystem and the vehicle terminal subsystem. The vehicle terminal system is used to acquire vehicle-side perception information and control vehicle movement based on vehicle driving decisions.