CN115214649B - Self-adaptive early warning method and system for driving control - Google Patents

Abstract

The invention discloses a self-adaptive early warning method for driving control, which is applied to a vehicle with a V2X vehicle-mounted unit, wherein a main vehicle can calculate whether collision risks exist among the main vehicle, a host lane, adjacent lanes and remote vehicles which are imported from a ramp in real time according to messages from a road side unit and other remote vehicles and by combining with the current driver state in the running process of a highway; if collision risk exists, alarm processing is carried out; and when detecting that the vehicle has a lane change intention, carrying out lane change risk detection, and prompting to cancel lane change operation when collision risk exists. The invention also discloses a corresponding system. By implementing the invention, the self-adaptive early warning of the vehicle running can be realized under various road shapes and weather conditions when the vehicle runs on the expressway, and the safety and the comfort of the vehicle running on the expressway are improved.

Description

Self-adaptive early warning method and system for driving control

Technical Field

The invention relates to the technical field of early warning of vehicles, in particular to a self-adaptive early warning method and system for driving control based on V2X road cooperation.

Background

At present, the driving mainly relies on sensors such as cameras and radars to sense the environment, but under severe environments such as vision shielding, rainy days and foggy days, the sensors can be inaccurately identified and even have functional failure; meanwhile, in a vehicle following part, such as an ACC function, most vehicles cannot well realize the ACC function under the condition of a curve, and in a lane keeping part, the lane line recognized by a camera is inaccurate, so that the lane transverse control and the lane matching are unsuccessful under the condition of bad weather possibly to cause the unsatisfactory lane keeping function effect; meanwhile, in the aspect of vehicle lane change control, most lane changes mainly consider that if the vehicle is detected to be safe at the beginning of lane change, lane change is carried out, and whether the vehicle has collision danger or not is monitored in real time in the lane change process is less considered; in addition, collision early warning in the aspect of straight roads is mostly considered aiming at whether collision danger exists between the vehicle lane changing process and surrounding vehicles, but the collision early warning in the aspect of lane changing of complex roads such as curves is less considered.

With the rapid development of the internet of vehicles technology V2X (Vehicle to Everything, internet of vehicles), the ability of vehicles to perceive the outside based on the V2X technology is becoming stronger. Because of the characteristics of high reliability and low delay of V2X, intelligent driving based on V2X is more and more paid attention. Compared with the traditional vehicle environment sensing schemes such as cameras and radars, the V2X is less influenced by environment changes, and can still work stably under severe environments such as vision shielding, rainy days and foggy days. With the development of the vehicle networking technology V2X, real-time perception and data interaction of people-vehicles-road-clouds based on the V2X technology are possible in some highways (such as highways and other closed areas), lane following and automatic lane changing functions based on the V2X technology are possible, and the vehicle can still stably work under severe environments such as vision shielding, rainy days and foggy days by utilizing the vehicle road cooperative control of the V2X technology, so that the research of the self-adaptive early warning based on the V2X vehicle road cooperative driving is of great significance.

Disclosure of Invention

The technical problem to be solved by the invention is to provide the self-adaptive early warning method and the self-adaptive early warning system for driving control, which can be used for predicting the collision risk in the driving process and have high safety and practicability.

In order to solve the above technical problem, as one aspect of the present invention, there is provided an adaptive early warning method for driving control, applied to a vehicle having a V2X on-board unit, comprising the steps of:

s10, the host vehicle receives the V2I message sent by the road side unit and the V2V message sent by the remote vehicle in the preset surrounding range in real time;

s11, determining the relative position relation between each far car and the own car according to the V2I message and each V2V message, determining the nearest far car of the own car in front of and behind Fang Juli of the same lane and adjacent lanes and the far car which is in cross running on a ramp as the current specific far car, obtaining the state information of each specific far car, and the safe distance, the distance value in unit time and the collision time sequence between the own car and each specific far car, and adjusting each safe distance according to the current state of a driver;

s12, determining whether collision risks exist between the specific far vehicles in front and behind and the specific far vehicles in cross running and the same lane according to the safety distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

S13, judging whether collision risks exist in front of or behind the corresponding adjacent lanes according to the safety distance, the distance value in unit time and the collision time sequence of each specific distant car on the adjacent lanes when judging that the lane changing intention exists; and reminding a driver to cancel the lane change when judging that the collision risk exists.

The V2I message sent by the road side unit is a local map message, and at least comprises road information, lane ID number and speed limit information; the V2V message includes vehicle status information of each vehicle within a predetermined distance range, the vehicle status information including: vehicle speed, vehicle position coordinates, steering wheel angle, heading angle, yaw angle, and acceleration information.

In the step S11, the step of determining the relative positional relationship between the remote vehicles and the host vehicle according to the V2I message and each V2V message further includes:

establishing a local coordinate system of the vehicle, taking the mass center of the vehicle as a coordinate origin, taking the running direction of the vehicle as a y axis, taking the direction of a driver pointing to the right hand as an x axis, and taking the direction of a vertical ground upwards as a z axis;

obtaining a difference value between the course angle of each far car and the course angle of the own car in a surrounding preset range, and judging the relative position between each far car and the own car according to the difference value, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

Performing coordinate translation transformation on each remote car, converting the coordinates of each remote car into a global coordinate system of the car, and calculating the relative distance between the remote car and the car;

determining and marking the specific position relation between each remote car and the host car by combining the difference value and the relative distance between the remote car and the host car; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross.

Wherein, the step S11 further comprises:

according to the state information of each specific remote car, carrying out iterative computation through a vector method to obtain a safety threshold value corresponding to each specific remote car and the own car in each time interval and a distance value in unit time; and stopping iterative computation of the corresponding specific remote car if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, and obtaining a collision time sequence of the specific remote car stopping the iteration.

Wherein, the step S11 further comprises:

the method comprises the steps of acquiring face and eye feature images of a driver by using a visual sensor above a front instrument of the driver, and identifying the current driving state behavior category and sight state of the driver by a mode identification method, wherein the driving state behavior category comprises: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

And according to the current driving state behavior category and the sight state, inquiring a prestored collision safety distance minimum value adjustment table to obtain corresponding adjustment values, and adjusting the safety threshold value corresponding to each specific remote car and the corresponding host car.

Wherein, the S12 further comprises:

when the distance value in unit time between the specific far car and the host car which are closest to each other on the same lane is judged to be smaller than or equal to a corresponding safety threshold value, determining that collision risk exists between the specific far car and the host car on the host lane, wherein the collision risk comprises the collision risk exists between the front Fang Teding far car and the host car, and the collision risk exists between the rear Fang Teding far car and the host car;

and determining that collision risk exists between the specific distant car and the host car in the cross running when the distance value in unit time between the specific distant car and the host car in the cross running is less than or equal to the corresponding safety threshold value.

Wherein, the S13 further comprises:

after detecting that the lane change intention exists in the vehicle, carrying out lane change detection on the adjacent lanes, and obtaining an optimal track for lane change of the adjacent lanes on the side and optimal lane change time by adopting optimal lane change track calculation fused by 5 times of polynomials and a genetic algorithm;

Judging that the distance value in unit time corresponding to each specific remote car of the adjacent channel on the side is larger than the corresponding safety threshold value, and judging that the risk of changing the channel to the adjacent channel on the side does not exist when no collision time sequence exists or the collision time sequence of the adjacent channel is larger than the optimal channel changing time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.

Correspondingly, in another aspect of the present invention, there is also provided an adaptive early warning system for driving control, applied to a vehicle having a V2X on-board unit, comprising:

the information receiving unit is used for receiving the V2I information sent by the road side unit in real time and receiving the V2V information sent by the remote vehicle in a preset surrounding range;

a specific far car state determining unit, configured to determine a relative position relationship between each far car and the host car according to the V2I message and each V2V message, determine a far car closest to the host car on the same lane, front and rear Fang Juli adjacent lanes, and a far car traveling on a ramp as a current specific far car, obtain state information of each specific far car, and obtain a safe distance between the host car and each specific far car, a distance value in unit time, and a collision time sequence, and adjust each safe distance according to a current state of a driver;

The driving risk identification processing unit is used for determining whether collision risks exist between the specific far vehicles in front and back directions on the same lane and the specific far vehicles in cross driving and the same vehicle according to the safety distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

the lane change risk identification processing unit is used for judging whether collision risks exist in front of or behind the corresponding adjacent lanes according to the safety distance, the distance value in unit time and the collision time sequence of each specific far car on the adjacent lanes when the lane change intention of the vehicle is judged; and reminding a driver to cancel the lane change when judging that the collision risk exists.

Wherein the specific remote car state determining unit further includes:

the relative position acquisition unit is used for establishing a local coordinate system of the vehicle, taking the mass center of the vehicle as a coordinate origin, the running direction of the vehicle as a y-axis, the direction of the right hand of a driver as an x-axis and the direction of the vertical ground upwards as a z-axis; and obtaining a difference value between the course angle of each remote car and the course angle of the host car in a surrounding preset range, and judging the relative position between each remote car and the host car according to the difference value, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

The relative distance acquisition unit is used for carrying out coordinate translation transformation on each far car, converting the coordinates of each far car into a global coordinate system of the car, and calculating the relative distance between the far car and the car;

the specific position relation acquisition unit is used for combining the difference value and the relative distance between the remote vehicle and the host vehicle, determining the specific position relation between each remote vehicle and the host vehicle and marking; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross;

the calculation processing unit is used for obtaining a safety threshold value corresponding to each specific remote car and the own car in each time interval and a distance value in unit time through iterative calculation by a vector method according to the state information of each specific remote car; if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, stopping iterative calculation of the corresponding specific remote car, and obtaining a collision time sequence of the specific remote car stopping the iteration;

the driving state recognition unit is used for acquiring face and eye feature images of a driver by utilizing a visual sensor above a front instrument of the driver and recognizing the current driving state behavior category and the sight state of the driver by a mode recognition method, wherein the driving state behavior category comprises the following steps: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

And the safety threshold value adjusting unit is used for inquiring a prestored collision safety distance minimum value adjusting table according to the current driving state behavior category and the sight state, obtaining corresponding adjusting values and adjusting the corresponding safety threshold value between each specific remote car and the own car.

Wherein the running risk identification processing unit further includes:

the lane risk judging unit is used for determining that collision risk exists between the specific far car and the host car on the same lane when the distance value in unit time between the specific far car which is closest to the same lane and the host car is smaller than or equal to the corresponding safety threshold value, wherein the collision risk comprises the collision risk of the front Fang Teding far car and the host car and the collision risk of the rear Fang Teding far car and the host car;

and the cross running risk judging unit is used for determining that collision risk exists between the specific far car and the car in the cross running when the distance value in unit time between the specific far car and the car in the cross running is judged to be smaller than or equal to the corresponding safety threshold value.

Wherein, the lane change risk identification processing unit further comprises:

the optimal track calculation unit is used for carrying out track change detection on the adjacent lanes after detecting that the lane change intention exists in the vehicle, and carrying out track change calculation on the adjacent lanes by adopting the optimal track change calculation fused by 5 times of polynomials and genetic algorithm to obtain the optimal track and the optimal track change time for the adjacent lanes on the side;

The lane change risk determining unit is used for determining that the lane change to the adjacent lane on the side is not at risk when the distance value in the unit time corresponding to each specific remote car on the adjacent lane on the side is larger than the corresponding safety threshold value and no collision time sequence exists or the collision time sequence of the collision time sequence is larger than the optimal lane change time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.

The embodiment of the invention has the following beneficial effects:

the invention provides a self-adaptive early warning method and a self-adaptive early warning system for driving control, wherein a host vehicle can calculate whether collision risks exist between the host vehicle and a far vehicle which is assembled in a lane, an adjacent lane and a ramp according to messages from a road side unit and other far vehicles and by combining with the current driver state in the running process of a highway; if collision risk exists, alarm processing is carried out; when detecting that the vehicle has a lane change intention, carrying out lane change risk detection, and prompting to cancel lane change operation when collision risk exists;

according to the invention, when the collision risk is calculated, the corresponding adjustment value is obtained according to the current driving state behavior category and the sight state, and the content of adjusting the corresponding safety threshold value between each specific remote car and the own car is added, so that the safety in the driving process can be further improved.

Meanwhile, the scheme provided by the invention can be suitable for various road-shaped working conditions and weather conditions, and has high safety and comfort.

Drawings

In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are required in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that it is within the scope of the invention to one skilled in the art to obtain other drawings from these drawings without inventive faculty.

Fig. 1 is a schematic flow chart of an embodiment of an adaptive early warning method for driving control according to the present invention;

FIG. 2 is a schematic view of the application scenario involved in FIG. 1;

FIG. 3 is a schematic diagram of the principle of classifying the relative positions between a host vehicle and a remote vehicle according to an embodiment of the present invention;

FIG. 4 is a vector coordinate diagram of the relative orientation of a remote vehicle with respect to a host vehicle in accordance with an embodiment of the present invention;

FIG. 5 is a schematic diagram showing more detailed relative position classification of a remote vehicle and a host vehicle in the same direction in an embodiment of the present invention;

FIG. 6 is a schematic diagram illustrating more detailed relative position classification when a remote vehicle is opposite to the host vehicle according to the embodiment of the present invention;

FIG. 7 is a schematic diagram of a remote vehicle at a left crossing position of the vehicle according to an embodiment of the present invention;

FIG. 8 is a schematic diagram of a remote vehicle at a right crossing position of the vehicle according to an embodiment of the present invention;

FIG. 9 is a vector analysis chart of collision risk calculation for a host vehicle and a remote vehicle traveling in a curve according to an embodiment of the present invention;

FIG. 10 is another vector analysis chart of collision risk calculation for a host vehicle and a remote vehicle traveling in a curve according to an embodiment of the present invention;

FIG. 11 is a schematic diagram of a lane change of a vehicle according to an embodiment of the present invention;

FIG. 12 is a more detailed flow chart of an adaptive early warning method for driving control according to an embodiment of the present invention;

FIG. 13 is a schematic diagram illustrating an embodiment of an adaptive early warning system for driving control according to the present invention;

FIG. 14 is a schematic diagram of the configuration of the specific remote status determination unit in FIG. 13;

fig. 15 is a schematic structural diagram of the running risk identification processing unit in fig. 13;

fig. 16 is a schematic structural diagram of the lane change risk identification processing unit in fig. 14.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings, for the purpose of making the objects, technical solutions and advantages of the present invention more apparent.

Referring to fig. 1, a schematic main flow chart of an embodiment of an adaptive early warning method for driving control according to the present invention is shown. As shown in fig. 2 to 12 together, the method of the present invention is applied to a vehicle having a V2X on-board unit (OBU), specifically, referring to an application environment schematic shown in fig. 2, in which a Road Side Unit (RSU) broadcasts surrounding road conditions and road information, traffic light information, map information, etc. to the vehicle through a V2I message; the vehicles communicate with each other through V2V messages, and mainly send or receive information such as the position, the speed and the direction of surrounding vehicles. In this embodiment, the method includes the steps of:

s10, the Host Vehicle (HV) receives the V2I message sent by the road side unit in real time and receives the V2V message sent by a Remote Vehicle (RV) within a preset range (such as 800 meters); more specifically, the V2I message sent by the road side unit is a local map message, which at least includes road information, a lane ID number, and speed limit information; the V2V message includes vehicle status information of each vehicle within a predetermined distance range, the vehicle status information including: vehicle speed (e.g., V), vehicle position coordinate position (X, Y, 0) (currently considered projected to the same plane), steering wheel angle (e.g., st_rv), head direction angle (e.g., H) _RV ) Heading angle (Heading), yaw angle, acceleration information, etc.; the host vehicle may be a vehicle that travels on a highway.

More specifically, as shown in fig. 2, there are a plurality of distant vehicles (RV) for the own vehicle (HV) and the surroundings ₁ 、RV ₂ 、RV ₃ 、RV ₄ 、RV ₅ 、RV ₆ 、RV ₇ ) Subsequent stepsThe relative position relation between each remote car and the own car needs to be determined, and meanwhile, whether collision risks exist between each remote car and the own car needs to be calculated.

S11, determining the relative position relation between each remote car and the own car according to the V2I message and each V2V message, determining the nearest remote car of the own car and the remote car which cross-runs on a ramp before and after the same lane and adjacent lanes as the current specific remote car, and obtaining the state information of each specific remote car, the safe distance between the own car and each specific remote car, the distance value in unit time and the collision time sequence;

in a specific example, in the step S11, it is necessary to first determine a relative positional relationship between the remote vehicles and the host vehicle, specifically, the method includes the steps of:

s110, firstly, a local coordinate system of the vehicle is established, as shown in FIG. 3, the centroid of the vehicle is taken as the origin of coordinates, the running direction of the vehicle is taken as the y axis, the direction of the right hand of the driver is taken as the x axis, and the direction of the vertical ground surface upwards is taken as the z axis;

S111, obtaining a difference value theta between the course angle of each remote car and the course angle of the host car in a surrounding preset range, and judging the relative position between each remote car and the host car according to the difference value theta, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

as shown in fig. 3, in the embodiment of the present invention, the boundary angles in each area division are determined at 30 °, 150 °, 210 ° and 330 °, the angles being calibrated in advance. More specifically, to prevent frequent jitter of the direction determination, the system may set a range of hysteresis angles, for RV that exceeds the boundary line and is within the range of hysteresis angles, not subject to classification variation, in the embodiment of the present invention, the hysteresis angle is calibrated to be 10 °, it being understood that this value is merely exemplary;

from the results, when θ is more than or equal to 0 and less than or equal to 30 degrees and 330 degrees, θ is more than or equal to 360 degrees, and the distant vehicle is in the same-direction running state of the vehicle; when theta is more than or equal to 30 degrees and less than 150 degrees, the far vehicle is in a left-crossing direction running state of the vehicle; when θ is more than or equal to 150 degrees and less than 210 degrees, the remote vehicle is in the opposite running state of the vehicle; when theta is more than or equal to 210 degrees and less than 330 degrees, the far vehicle is in a right-crossing direction running state of the vehicle.

S112, carrying out coordinate translation transformation on each remote car, converting the coordinates of each remote car into a global coordinate system of the car, and calculating the relative distance between the remote car and the car;

More specifically, as shown in fig. 4 to 8, the relative front-rear positional relationship of the host vehicle and the remote vehicle is determined by conversion and calculation of the coordinate system. Fig. 4 shows a relationship diagram of the remote vehicle RV in the main vehicle HV coordinate system. The remote vehicle RV coordinates are converted to global coordinates of the host vehicle HV, where α is the angle of rotation of the HV vehicle y axis, and here may be considered as the heading angle of the remote vehicle RV and the host vehicle HV, with the clockwise direction being the positive direction.

x _RV>HV ＝(X _RV -X _HV )cosα-(Y _RV -Y _HV )sinα

y _RV>HV ＝(X _RV -X _HV )sinα-(Y _RV -Y _HV )cosα

Wherein X is _HV 、X _RV The abscissa representing HV and RV in global coordinates; y is Y _HV 、Y _RV Representing the ordinate of HV and RV in global coordinates; x is x _RV>HV Represents the abscissa of RV relative to HV coordinates; y is _RV>HV Indicating the ordinate of RV with respect to HV coordinates.

When y is _RV>HV > 0, indicating that RV is in front of HV; when y is _RV>HV < 0, indicating that RV is aft of HV.

S113, determining and marking the specific position relation between each remote car and the host car by combining the difference value theta and the relative distance between the remote car and the host car; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross.

More specifically, as shown in fig. 5, when the far vehicle is running in the same direction, the positions of surrounding vehicles are classified into 10 different areas by taking the head position of the vehicle as the center, and the classification method of the areas is that the width of a Chinese lane is 2.75-3.5 meters, and the lane width is 3.5 meters. Assuming that the vehicle runs in the middle of a lane, taking the abscissa of the position of the vehicle head as the reference:

Table 1 below shows a relative position classification list of RV versus HV at the same direction:

TABLE 1

Coordinate interval	Position definition of RV relative to HV
		-1.75≤x RV>HV ≤1.75&&y RV>HV ≥0	Front (Ahead)
1.75≤x RV>HV <5.25&&y RV>HV ≥0	Right front (Ahead Right)
		x RV>HV >5.25&&y RV>HV ≥0	Far Right front (Ahead Far Right)
-5.25≤x RV>HV ≤-1.75&&y RV>HV ≥0	Left front (Ahead Left)
		x RV>HV ≤-5.25&&y RV>HV ≥0	Far Left front (Ahead Far Left)
-1.75≤x RV>HV ≤1.75&&y RV>HV <0	Rear (Behind)
		-5.25<x RV>HV ≤-1.75&&y RV>HV <0	Left rear (Behide Left)
x RV>HV <-5.25&&y RV>HV <0	Far Left rear (Behind Far Left)
		1.75≤x RV>HV <5.25&&y RV>HV <0	Right rear (Behide Right)
x RV>HV >5.25&&y RV>HV <0	Far Right rear (Behide Far Right)

As shown in fig. 6, when the remote vehicle is traveling in the opposite direction, the relative position of the remote vehicle is determined centering on the head position of the host vehicle. Table 2 below shows a relative position classification list of the subtended RV with respect to HV:

TABLE 2

Coordinate interval	RV relative HV orientation
		-1.75≤x RV>HV ≤1.75&&y RV>HV ≥0	Front direction (Oncoming)
1.75≤x RV>HV <5.25&&y RV>HV ≥0	Right front direction (Oncoming right)
		x RV>HV >5.25&&y RV>HV ≥0	Far right front facing (Oncoming far right)
-5.25≤x RV>HV ≤-1.75&&y RV>HV ≥0	Left front facing (Oncoming left)
		x RV>HV ≤-5.25&&y RV>HV ≥0	Far left front facing (Oncoming far left)

While the direction of travel of the vehicle at the intersection (including the entrance of the vehicle at the expressway ramp) is 30 DEG<When θ is less than or equal to 150 °, the distant vehicle RV is in the left crossing direction of the vehicle HV, and RV is in the left x of HV _RV>HV Less than or equal to 0, judging that RV is in the direction of the left intersection (Intersection Left) of HV for coming vehicles, as shown in FIG. 7;

when 210 DEG<When θ is less than or equal to 330 degrees, the distant vehicle RV is in the right crossing direction of the vehicle HV, and simultaneously represents the right ramp afflux condition of the expressway, and RV is in the right x of HV _RV>HV And (5) 0, judging that RV is in the right crossing (Intersection Right) direction of HV to come, as shown in FIG. 8.

From the above analysis, the direction and position of the far car RV with respect to the own car HV falls into 18 position areas as shown in table 3 below; for remote vehicles that fail to locate a classification, they should be classified as unclassified (None).

TABLE 3 Table 3

By combining the tables, according to the lane information sent by the RSU, RV vehicle sequences of the same lane and adjacent lanes of the HV can be screened, nearest distant vehicles of the vehicle on the same lane or adjacent lanes in front and behind Fang Juli and distant vehicles of the vehicle on the ramp can be identified, the vehicles are determined to be specific distant vehicles of the vehicle, and the classification serial number of the position area of each distant vehicle is marked and recorded by adopting a flag.

It is understood that S11 further includes:

s114, carrying out iterative computation through a vector method according to the state information of each specific remote car to obtain a safety threshold d corresponding to each specific remote car and the host car in each time interval _w,n Distance value per unit time DCPA _n The method comprises the steps of carrying out a first treatment on the surface of the And stopping iterative calculation of the corresponding specific remote car if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, and obtaining a collision time sequence TCPA of the specific remote car stopping the iteration.

The step S114 is implemented by adopting a vector-based vehicle 360-degree recognition and vehicle collision pre-warning algorithm, and specifically, as shown in fig. 9, the speeds of the far vehicle RV and the host vehicle in curve driving are V respectively _HV 、V _RV Steering wheel angle St _HV 、St _RV The direction angle of the head of the vehicle is H _HV The method comprises the steps of carrying out a first treatment on the surface of the Wherein H is _RV Specifically, the head direction angle is positive anticlockwise by the included angle between the head advancing direction and the Y axis of the geodetic coordinate system; alpha _HV 、α _RV The steering angle is HV and RV, the steering angle is positive in the clockwise direction, and the steering angle is negative in the anticlockwise direction; vector vehicle speedAt B ₁ As a starting point, alpha _HV Rotation, RV is +.about.HV traveling vehicle speed>At->Projection of (2) is asWherein A is ₁ Is a projection point; the purpose is to find the distance of RV from HV in unit time (i.e. DCPA _n )。

Wherein the projectionThe calculation formula of (2) is as follows:

wherein θ is ₁ Is a vectorAnd->Is included in the plane of the first part;

here, the

Thus (2)

Then

When n=1, the velocity of RV relative to HV isTo find the closest distance of HV to RV then a typical mathematical problem is then formed: one point outside line segment HV to line segment->Is the shortest distance of:

point-of-nothing HV line sectionIn which position equation (4) holds, a coefficient can be setThen:

the physical meaning of the expression is as follows: if A ₁ At vector On the other hand, this point is the point of the shortest distance of RV with respect to HV (CPA when n=1 in the first cycle ₁ ) Vector->I.e. DCPA ₁ The method comprises the steps of carrying out a first treatment on the surface of the If A ₁ The point is->On the extension line of (2) can be used->Representation of DCPA ₁ The method comprises the steps of carrying out a first treatment on the surface of the If A ₁ At->On the extension line of (2), then +.>Representation of DCPA ₁ 。

Fig. 10 shows that when HV and RV are traveling in a curve, RV is n=3 relative to HV, and CPA is obtained ₃ WhereinThe vector +.>And->Equal in size and opposite in direction; thus when n=1, B ₁ And P ₁ Coordinates have B ₁ Point coordinates:

then P ₁ Coordinates:

when n=n,

B _n the points are based on coordinates of a GPS coordinate system (global coordinate system):

wherein:

velocity and acceleration vectors for the initial state RV; alpha _RV,0 For the beginningThe steering angle in the initial state is RV wheel steering angle because the national standard of the V2X application layer prescribes that the steering wheel angle St can be obtained from the whole bus>i _RV Is the steering gear ratio of the RV.

P _n The points are based on coordinates of a GPS coordinate system (global coordinate system):

wherein:

velocity and acceleration vectors for the initial state HV; alpha _HV 0 is the steering angle in the initial state, and since the V2X application layer national standard prescribes that the steering angle St can be obtained from the whole bus, the HV wheel steering angle +. >i _HV Is the steering gear ratio of the RV.

Vector method based on B _n Point, P _n The analysis of the points is irrelevant to the type of path of the vehicle, so that real-time calculation of whether the HV vehicle has collision risk to surrounding vehicles can be performed according to different directions of the HV in which the RV is positioned and the local map information sent by the RSU. As shown in FIG. 11, HV and surrounding RV need to be considered during lane change ₁ 、RV ₂ 、RV ₃ 、RV ₄ 、RV ₅ 、RV ₆ If there is a risk of collision, the HV can change lanes only if there is no risk of collision between the HV and surrounding vehicles.

Therefore, the safety distance model of HV and RV of the invention is as follows:

when V is _RV >At 0 time

When V is _RV When=0

Wherein V is _rel For HV and RV relative vehicle speed, R _min Is the minimum safe distance.

For HV and surrounding RV ₁ 、RV ₂ 、RV ₃ 、RV ₄ 、RV ₅ 、RV ₆ Calculation of whether there is a collision risk, in order to be located in front of HV and in the RV of the same lane ₁ For example, as can be seen from fig. 5, when the sum of RV and HV is n=1, the sum isn=2, isn=3 is +.>Will->Respectively projected to vector +.>Applying; due toVehicle speed vector>Equal in size and opposite in direction; thus, in each time interval Δt (set Δt=1s), at n=n(n can be actually calibrated) when:

when V is _RV >At the time of 0, the temperature of the liquid,

when V is _RV When the value of the sum is =0,

wherein,is->And->Is included in the plane of the first part; / >Is->Andis included in the plane of the first part; />Is->And->Included angle R _min Is the minimum safe distance;

in the vector calculation, if DCPA is used in n calculation times _n ≤d _w,n Stopping the calculationAt this time, HV can be derived in the future T _warning After time there is a risk of a forward collision. Wherein it can be seen from the formulas (5) (6) that at the nth calculation time, it is possible to obtain:

from the opposite forward RV ₁ As can be seen from the collision time vector calculation, for RV ₂ 、RV ₃ 、RV ₄ 、RV ₅ 、RV ₆ The collision time vector calculation method is the same as that of HV.

Still further, the step S11 further includes:

s115, acquiring face and eye feature images of a driver by using a vision sensor above a front instrument of the driver, and identifying the current driving state behavior category and the sight state of the driver by a mode identification method, wherein the driving state behavior category comprises: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

s116, inquiring a prestored collision safety distance minimum value adjustment table according to the current driving state behavior category and the sight state, obtaining corresponding adjustment values, and adjusting safety thresholds corresponding to specific remote vehicles and the vehicle.

More specifically, in one example, a visual sensor (camera) above a front instrument of the driver may be utilized to obtain face and eye feature images of the driver, real-time monitoring of the state and the line of sight of the driver is achieved through a pattern recognition method (such as a convolutional neural network model), and the driving state behaviors of the driver are divided into: normal driving (l=1), fatigue driving (l=2), distraction driving (l=3), call answering (l=4), smoking (l=5), emotional agitation driving (l=6), drunk driving (l=7) 7 modes; by judging the driver's sight line, the driver's sight line is divided into 3 sight line states of a central control area (q=1), a front windshield area (q=2) and an outside rear view mirror area (q=3), and the minimum value R of the collision safety distance is adjusted in real time _min (specific value R _min Derived from calibration), specific adjustment values are shown in table 4 below:

TABLE 4R under different conditions _min List of adjustment values

	L＝1	L＝2	L＝3	L＝4	L＝5	L＝6	L＝7
								q＝1	3	5	7	10	12	13	15
q＝2	3	5	7	10	14	15	18
								q＝3	4	5	7	10	16	18	25

S12, determining whether collision risks exist between the specific far vehicles in front and behind and the specific far vehicles running in a cross way on the same lane and the same vehicle according to the safety distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

In a specific example, the step S12 further includes:

judgingWhen the distance value in unit time between the specific distant car closest to the same lane and the host car is smaller than or equal to the corresponding safety threshold value (i.e. DCPA _n ≤d _w,n ) Determining that collision risks exist between the specific far vehicle and the vehicle on the lane, wherein the collision risks include the collision risks of the front Fang Teding far vehicle and the collision risks of the rear Fang Teding far vehicle and the vehicle;

more specifically, by judging DCPA _n ≤d _w,n &&(RV and HV are in the same lane)&&(flag= =1|2|4), if the condition is satisfied, judging that there is a collision risk in front of HV in the same lane; DCPA _n ≤d _w,n &&(RV and HV are in the same lane)&&(flag= 6 7 9) if the condition is satisfied, judging that there is a collision risk behind the HV same lane;

and determining that collision risk exists between the specific distant car and the host car in the cross running when the distance value in unit time between the specific distant car and the host car in the cross running is less than or equal to the corresponding safety threshold value.

More specifically, DCPA _n ≤d _w,n &&(flag= =16||17) if the condition is satisfied, judging that there is a collision risk of the HV and the ramp converging into the RV vehicle;

s13, judging whether collision risks exist in front of or behind the corresponding adjacent lanes according to the safety distance, the distance value in unit time and the collision time sequence of each specific distant car on the adjacent lanes when judging that the lane change effect diagram exists on the own car; and reminding a driver to cancel the lane change when judging that the collision risk exists.

In a specific example, the step S13 further includes:

s130, after detecting that the lane change intention exists in the vehicle, carrying out lane change detection on the adjacent lanes, and obtaining an optimal track for lane change of the adjacent lanes on the side and optimal lane change time by adopting optimal lane change track calculation fused by 5 times of polynomials and a genetic algorithm; in a specific example, if it is detected that the angle of turning the steering wheel exceeds a certain angle or turning on the turn signal lamp, it is determined that the host vehicle has a lane change intention;

s131, judging that the distance value in unit time corresponding to each specific remote car of the adjacent channel on the side is larger than a corresponding safety threshold value, and judging that the risk of changing the channel to the adjacent channel on the side does not exist when no collision time sequence exists or the collision time sequence of the adjacent channel is larger than the optimal channel changing time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.

Specifically, if DCPA is judged _n ≤d _w,n &&(RV and HV in adjacent lanes)&&(flag= =1|2|4), if the collision risk exists in front of the HV adjacent lane, the collision risk exists in front of the HV left adjacent lane or the HV right adjacent lane is sent to the driver, and the driver is reminded to cancel lane changing; if judging to DCPA _n ≤d _w,n &&(RV and HV in adjacent lanes) &&(flag= =6|7|9), then indicate that there is a collision risk behind the HV adjacent lane, then send out to the driver that there is a collision risk behind HV left adjacent lane or right adjacent lane, remind the driver to cancel lane change.

The following describes the principle of the optimal lane change trajectory planning of the fusion of the 5 th order polynomial and the genetic algorithm in S130 of the present invention:

the adoption of the high order polynomial as the ideal reference track change track has the following two advantages: firstly, the curve of the high-order polynomial track is smoother, the first derivative and the second derivative of the curve are continuous smooth functions, and abrupt change does not occur in the control process; secondly, the high-frequency component of the high-order polynomial track curve is smaller, and the system is easy to carry out compensation control on the high-frequency component through a feedback system;

the 5 th order polynomial lane change track expression used in the invention is:

longitudinal displacement x (t) and transverse displacement y (t) are each a function of time t

Wherein a is ₀ ～a ₅ Coefficient to be determined for longitudinal displacement trajectory, b ₀ ～b ₅ Coefficient initial state to be determined for transverse displacement trackAnd target state->If the state is known, then the undetermined coefficients of equation (21) can be solved.

At the initial and final moments of the lane change, the running state of the automobile tends to be stable, acceleration is not generated, and transverse speed is not generated, so that the kinematic characteristic of the automobile can be met. The initial state can be expressed as Wherein v is _xin Representing the longitudinal initial state speed.

The target state may be expressed asWherein v is _xfin The longitudinal speed after the lane change is finished is represented, L is the longitudinal displacement in the lane change process, h represents the lane width and is generally 3.75m. When t=0 is set, lane change is started and the vehicle mass center is located at the origin of coordinates at the moment, and t=t ₀ And (3) finishing channel changing, and substituting the channel changing into a formula (21) to obtain:

solving second and third derivatives of the formula (21) to obtain the transverse and longitudinal acceleration of the lane change track based on the fifth order polynomial:

substituting equations (22) and (23) into equation (24) to obtain an extremum value:

wherein a is _mx Is the maximum value of the track longitudinal acceleration; a, a _my Is the maximum of the lateral acceleration of the track. Since the lane width h is known, v _xin CAN be obtained through the CAN bus of the HV vehicle, so that the extreme value of the acceleration is only equal to v _xfin L and lane change time t ₀ And (5) correlation. Therefore, lane change track evaluation index J:

wherein a is _max Representing the maximum acceleration of the vehicle, L _max To change the maximum longitudinal displacement, t _cmax Is the maximum value of the channel changing time; w (w) ₁ 、w ₂ 、w ₃ The relation of the weight coefficient and the three is as follows: w (w) ₁ +w ₂ +w ₃ =1. L represents longitudinal displacement of lane change, and represents influence on traffic flow, wherein the smaller the value is, the smaller the influence is; t is t _c The channel change time of the channel change is represented, and the channel change efficiency is represented as the value is smaller, the channel change efficiency is higher.

The problem of lane change trajectory optimization based on genetic algorithm is therefore described as:

the fitness function is brought into a genetic algorithm module to carry out iterative optimization, and v corresponding to the J minimum condition can be obtained _xfin L and t _c Thus, the complete boundary condition is obtained, each parameter of the 5 th order polynomial can be obtained, and finally the optimal track change track is obtained.

For more details of the steps of the method provided by the present invention reference is made to the contents of fig. 12.

Referring to fig. 13, a schematic structural diagram of an embodiment of an adaptive early warning system for driving control according to the present invention is shown. As shown in fig. 14 to 16 together, in the present embodiment, the adaptive early warning system for driving control is applied to a vehicle having a V2X on-board unit, and more specifically, the adaptive early warning system 1 for driving control includes:

the information receiving unit 10 is used for receiving the V2I message sent by the road side unit in real time and receiving the V2V message sent by the far vehicle in the preset range around when the vehicle runs on the expressway;

a specific far car state determining unit 11, configured to determine a relative positional relationship between each far car and the host car according to the V2I message and each V2V message, determine a far car closest to the host car in front of and behind Fang Juli the same lane and adjacent lanes, and a far car traveling across a ramp as a current specific far car, obtain state information of each specific far car, and obtain a safe distance between the host car and each specific far car, a distance value in unit time, and a collision time sequence, and adjust each safe distance according to a current state of a driver;

A driving risk recognition processing unit 12, configured to determine whether there is a collision risk between the specific far vehicle and the host vehicle that travel in a cross manner, and the specific far vehicle that travel in front and back on the same host lane, according to the safe distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

the lane change risk identification processing unit 13 is configured to determine whether a collision risk exists in front of or behind the corresponding adjacent lane according to the safety distance, the distance value in unit time, and the collision time sequence of each specific distant vehicle on the adjacent lane when it is determined that the lane change effect diagram exists on the host vehicle; and reminding a driver to cancel the lane change when judging that the collision risk exists.

The V2I message sent by the road side unit is a local map message, and at least comprises road information, lane ID number and speed limit information; the V2V message includes vehicle status information of each vehicle within a predetermined distance range, the vehicle status information including: vehicle speed, vehicle position coordinates, steering wheel angle, heading angle, yaw angle, and acceleration information.

Wherein the specific remote car state determining unit 11 further includes:

a relative position obtaining unit 110, configured to establish a local coordinate system of the vehicle, with a centroid of the vehicle as an origin of coordinates, a vehicle running direction as a y-axis, a direction of a driver pointing to a right hand as an x-axis, and a vertical ground direction as a z-axis direction; and obtaining a difference value theta between the course angle of each remote car and the course angle of the host car in a surrounding preset range, and judging the relative position between each remote car and the host car according to the difference value theta, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

A relative distance obtaining unit 111, configured to perform coordinate translation transformation on each remote vehicle, convert the coordinates of each remote vehicle into a global coordinate system of the vehicle, and calculate a relative distance between the remote vehicle and the vehicle;

a specific positional relationship obtaining unit 112, configured to determine and mark specific positional relationships between the remote vehicles and the host vehicle by combining the difference θ and the relative distance between the remote vehicles and the host vehicle; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross;

the computing unit 113 is configured to obtain a safety threshold value corresponding to each specific remote car and the own car in each time interval and a distance value in unit time by performing iterative computation according to the state information of each specific remote car through a vector method; if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, stopping iterative calculation of the corresponding specific remote car, and obtaining a collision time sequence of the specific remote car stopping the iteration;

the driving state identifying unit 114 is configured to obtain a face and eye feature image of a driver by using a vision sensor above a front meter of the driver, and identify a current driving state behavior category of the driver and a sight line state by using a pattern identifying method, where the driving state behavior category includes: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

The safety threshold value adjusting unit 115 is configured to query a prestored collision safety distance minimum value adjusting table according to the current driving state behavior category and the sight state, obtain a corresponding adjusting value, and adjust the safety threshold value corresponding to each specific remote vehicle and the host vehicle.

Wherein the running risk identification processing unit 12 further includes:

the host lane risk judging unit 120 is configured to determine that a collision risk exists between the host vehicle and the specific far vehicle on the same lane when the distance value in unit time between the host vehicle and the specific far vehicle closest to the same lane is less than or equal to the corresponding safety threshold, where the collision risk includes that the front Fang Teding far vehicle and the host vehicle exist, and that the rear Fang Teding far vehicle and the host vehicle exist;

the cross travel risk judging unit 121 is configured to determine that there is a collision risk between the specific far car and the own car in the cross travel when it is determined that the distance value in unit time between the specific far car and the own car in the cross travel is less than or equal to the corresponding safety threshold.

Wherein the lane change risk identification processing unit 13 further includes:

the optimal track calculation unit 130 is configured to perform lane change detection on the adjacent lane after detecting that the lane change intention exists in the host vehicle, and obtain an optimal track and an optimal lane change time for lane change of the adjacent lane on the host side by adopting optimal lane change track calculation fused by a polynomial and a genetic algorithm for 5 times;

The lane change risk determining unit 131 is configured to determine that there is no risk of lane change to the neighboring lane when it is determined that the distance value in the unit time corresponding to each specific remote vehicle of the neighboring lane is greater than the corresponding safety threshold, and no collision time sequence exists or the collision time sequences thereof are both greater than the optimal lane change time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.

For more details, reference is made to the previous descriptions of fig. 1 to 11, which are not described in detail here.

The embodiment of the invention has the following beneficial effects:

the invention provides a self-adaptive early warning method and a self-adaptive early warning system for driving control, wherein a host vehicle can calculate whether collision risks exist between the host vehicle and a far vehicle which is assembled in a lane, an adjacent lane and a ramp according to messages from a road side unit and other far vehicles and by combining with the current driver state in the running process of a highway; if collision risk exists, alarm processing is carried out; when detecting that the vehicle has a lane change intention, carrying out lane change risk detection, and prompting to cancel lane change operation when collision risk exists;

according to the invention, when the collision risk is calculated, the corresponding adjustment value is obtained according to the current driving state behavior category and the sight state, and the content of adjusting the corresponding safety threshold value between each specific remote car and the own car is added, so that the safety in the driving process can be further improved.

Meanwhile, the scheme provided by the invention can be suitable for various road-shaped working conditions and weather conditions, and has high safety and comfort.

It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above disclosure is only a preferred embodiment of the present invention, and it is needless to say that the scope of the invention is not limited thereto, and therefore, the equivalent changes according to the claims of the present invention still fall within the scope of the present invention.

Claims (11)

1. An adaptive early warning method for driving control, comprising:

s10, the host vehicle receives the V2I message sent by the road side unit and the V2V message sent by the remote vehicle in the preset surrounding range in real time;

s11, determining the relative position relation between each far car and the own car according to the V2I message and each V2V message, determining the nearest far car of the own car in front of and behind Fang Juli of the same lane and adjacent lanes and the far car which is in cross running on a ramp as the current specific far car, obtaining the state information of each specific far car, and the safe distance, the distance value in unit time and the collision time sequence between the own car and each specific far car, and adjusting each safe distance according to the current state of a driver;

s12, determining whether collision risks exist among the specific far vehicles in front and behind on the same lane, the specific far vehicles in cross running and the vehicle according to the safety distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

S13, judging whether collision risks exist in front of or behind the corresponding adjacent lanes according to the safety distance, the distance value in unit time and the collision time sequence of each specific distant car on the adjacent lanes when judging that the lane changing intention exists; and reminding a driver to cancel the lane change when judging that the collision risk exists.

2. The method of claim 1, wherein the V2I message sent by the roadside unit is a local map message including at least road information, lane ID number, speed limit information; the V2V message includes vehicle status information of each vehicle within a predetermined distance range, the vehicle status information including: vehicle speed, vehicle position coordinates, steering wheel angle, heading angle, yaw angle, and acceleration information.

3. The method of claim 2, wherein in S11, the step of determining the relative positional relationship between the remote vehicles and the host vehicle from the V2I message and each V2V message further comprises:

establishing a local coordinate system of the vehicle, taking the mass center of the vehicle as a coordinate origin, taking the running direction of the vehicle as a y axis, taking the direction of a driver pointing to the right hand as an x axis, and taking the direction of a vertical ground upwards as a z axis;

Obtaining a difference value between the course angle of each far car and the course angle of the own car in a surrounding preset range, and judging the relative position between each far car and the own car according to the difference value, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

performing coordinate translation transformation on each remote car, converting the coordinates of each remote car into a global coordinate system of the car, and calculating the relative distance between the remote car and the car;

determining and marking the specific position relation between each remote car and the host car by combining the difference value and the relative distance between the remote car and the host car; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross.

4. The method of claim 3, wherein S11 further comprises:

according to the state information of each specific remote car, carrying out iterative computation through a vector method to obtain a safety threshold value corresponding to each specific remote car and the own car in each time interval and a distance value in unit time; and stopping iterative computation of the corresponding specific remote car if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, and obtaining a collision time sequence of the specific remote car stopping the iteration.

5. The method of claim 4, wherein S11 further comprises:

the method comprises the steps of acquiring face and eye feature images of a driver by using a visual sensor above a front instrument of the driver, and identifying the current driving state behavior category and sight state of the driver by a mode identification method, wherein the driving state behavior category comprises: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

and according to the current driving state behavior category and the sight state, inquiring a prestored collision safety distance minimum value adjustment table to obtain corresponding adjustment values, and adjusting the safety threshold value corresponding to each specific remote car and the corresponding host car.

6. The method of claim 5, wherein S12 further comprises:

when the distance value in unit time between the specific far car and the host car which are closest to each other on the same lane is judged to be smaller than or equal to a corresponding safety threshold value, determining that collision risk exists between the specific far car and the host car on the host lane, wherein the collision risk comprises the collision risk exists between the front Fang Teding far car and the host car, and the collision risk exists between the rear Fang Teding far car and the host car;

And determining that collision risk exists between the specific distant car and the host car in the cross running when the distance value in unit time between the specific distant car and the host car in the cross running is less than or equal to the corresponding safety threshold value.

7. The method of claim 4, wherein S13 further comprises:

after detecting that the lane change intention exists in the vehicle, carrying out lane change detection on the adjacent lanes, and obtaining an optimal track for lane change of the adjacent lanes on the side and optimal lane change time by adopting optimal lane change track calculation fused by 5 times of polynomials and a genetic algorithm;

judging that the distance value in unit time corresponding to each specific remote car of the adjacent channel on the side is larger than the corresponding safety threshold value, and judging that the risk of changing the channel to the adjacent channel on the side does not exist when no collision time sequence exists or the collision time sequence of the adjacent channel is larger than the optimal channel changing time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.

8. An adaptive early warning system for driving control, comprising:

the information receiving unit is used for receiving the V2I information sent by the road side unit in real time and receiving the V2V information sent by the remote vehicle in a preset surrounding range;

A specific far car state determining unit, configured to determine a relative position relationship between each far car and the host car according to the V2I message and each V2V message, determine a far car closest to the host car on the same lane, front and rear Fang Juli adjacent lanes, and a far car traveling on a ramp as a current specific far car, obtain state information of each specific far car, and obtain a safe distance between the host car and each specific far car, a distance value in unit time, and a collision time sequence, and adjust each safe distance according to a current state of a driver;

the driving risk identification processing unit is used for determining whether collision risks exist between the specific far vehicles in front and behind on the same lane and the specific far vehicles in cross driving and the host vehicle according to the safety distance and the distance value in unit time; if collision risk exists, alarming and prompting the driver;

the lane change risk identification processing unit is used for judging whether collision risks exist in front of or behind the corresponding adjacent lanes according to the safety distance, the distance value in unit time and the collision time sequence of each specific far car on the adjacent lanes when the lane change intention of the vehicle is judged; and reminding a driver to cancel the lane change when judging that the collision risk exists.

9. The system of claim 8, wherein the specific remote vehicle status determination unit further comprises:

the relative position acquisition unit is used for establishing a local coordinate system of the vehicle, taking the mass center of the vehicle as a coordinate origin, the running direction of the vehicle as a y-axis, the direction of the right hand of a driver as an x-axis and the direction of the vertical ground upwards as a z-axis; and obtaining a difference value between the course angle of each remote car and the course angle of the host car in a surrounding preset range, and judging the relative position between each remote car and the host car according to the difference value, wherein the relative position comprises the following steps: the same direction, opposite direction and cross direction;

the relative distance acquisition unit is used for carrying out coordinate translation transformation on each far car, converting the coordinates of each far car into a global coordinate system of the car, and calculating the relative distance between the far car and the car;

the specific position relation acquisition unit is used for combining the difference value and the relative distance between the remote vehicle and the host vehicle, determining the specific position relation between each remote vehicle and the host vehicle and marking; the specific positional relationship at least comprises: front, right front, left front, rear, left rear, right rear, front facing, right front facing, left cross, and right cross;

The calculation processing unit is used for obtaining a safety threshold value corresponding to each specific remote car and the own car in each time interval and a distance value in unit time through iterative calculation by a vector method according to the state information of each specific remote car; if the distance value in the unit time is smaller than or equal to the corresponding safety threshold value, stopping iterative calculation of the corresponding specific remote car, and obtaining a collision time sequence of the specific remote car stopping the iteration;

the driving state recognition unit is used for acquiring face and eye feature images of a driver by utilizing a visual sensor above a front instrument of the driver and recognizing the current driving state behavior category and the sight state of the driver by a mode recognition method, wherein the driving state behavior category comprises the following steps: normal driving, fatigue driving, distraction driving, call answering, smoking, emotion exciting driving and drunk driving; the line-of-sight state includes: in the central control area, in the front windscreen area and in the outside mirror area;

and the safety threshold value adjusting unit is used for inquiring a prestored collision safety distance minimum value adjusting table according to the current driving state behavior category and the sight state, obtaining corresponding adjusting values and adjusting the corresponding safety threshold value between each specific remote car and the own car.

10. The system of claim 9, wherein the driving risk identification processing unit further comprises:

the lane risk judging unit is used for determining that collision risk exists between the specific far car and the host car on the same lane when the distance value in unit time between the specific far car which is closest to the same lane and the host car is smaller than or equal to the corresponding safety threshold value, wherein the collision risk comprises the collision risk of the front Fang Teding far car and the host car and the collision risk of the rear Fang Teding far car and the host car;

and the cross running risk judging unit is used for determining that collision risk exists between the specific far car and the car in the cross running when the distance value in unit time between the specific far car and the car in the cross running is judged to be smaller than or equal to the corresponding safety threshold value.

11. The system of claim 10, wherein the lane change risk identification processing unit further comprises:

the optimal track calculation unit is used for carrying out track change detection on the adjacent lanes after detecting that the lane change intention exists in the vehicle, and carrying out track change calculation on the adjacent lanes by adopting the optimal track change calculation fused by 5 times of polynomials and genetic algorithm to obtain the optimal track and the optimal track change time for the adjacent lanes on the side;

The lane change risk determining unit is used for determining that the lane change to the adjacent lane on the side is not at risk when the distance value in the unit time corresponding to each specific remote car on the adjacent lane on the side is larger than the corresponding safety threshold value and no collision time sequence exists or the collision time sequence of the collision time sequence is larger than the optimal lane change time; otherwise, judging that the risk exists for the lane change of the adjacent lane on the side.