CN114475573A

CN114475573A - Fluctuating road condition identification and vehicle control method based on V2X and vision fusion

Info

Publication number: CN114475573A
Application number: CN202210134605.4A
Authority: CN
Inventors: 史兴领; 陈英箭; 檀杰; 陈辰; 王磊
Original assignee: Kedahuochuang Hefei Intelligent Automobile Technology Co ltd
Current assignee: Kedahuochuang Hefei Intelligent Automobile Technology Co ltd
Priority date: 2022-02-14
Filing date: 2022-02-14
Publication date: 2022-05-13
Anticipated expiration: 2042-02-14
Also published as: CN114475573B

Abstract

The invention discloses a fluctuating road condition identification and vehicle control method based on V2X and vision fusion, which relates to the technical field of intelligent driving, comprises five groups of signal links, and specifically comprises the following steps: acquiring original pavement data of each signal link; processing data obtained by sensing and detecting the V2I, the V2V and the road surface fluctuation based on vision by adopting a direct detection method to sequentially obtain an international flatness index (IRI)₁、IRI₂And IRI₃(ii) a To the front-based car lightsProcessing the vertical position jump data by adopting a response type detection method to obtain an international flatness index IRI₄(ii) a Weighting the international flatness indexes acquired from different sources to obtain the final international flatness index IRI_f(ii) a Binding to IRI_fAnd the international flatness index evaluation index optimizes and adjusts the current running speed or running track, and improves the safety and comfort of running on the rough road.

Description

Fluctuating road condition identification and vehicle control method based on V2X and vision fusion

Technical Field

The invention relates to the technical field of intelligent driving, in particular to a rough road condition identification and vehicle control method based on V2X and vision fusion.

Background

With the evolution and development of intelligent driving technology, the requirement of high-level intelligent driving on the whole scene coverage is more and more strong. On a road section where heavy vehicles pass through more roads or the roads are not maintained timely, a steep and deep pit is common, and in some suburb scenes, a deceleration strip made of nonstandard cement and the like or a road and bridge joint is common, and a deep pit is also common. The rough road surfaces bring great threat and influence to driving safety and comfort, especially in the dark night when the pits are filled with rainwater, on the premise of lacking effective sensing means, uncomfortable rough road surfaces are easily caused, and damage to a vehicle chassis and tires can be caused under extreme working conditions;

the current technologies for dealing with road undulations are mainly classified into two types: one is a suspension adjustment technique based on passive road condition perception; however, due to the passive sensing characteristic, the system can be actuated only when contacting with a rough road surface, the response has lag, and under some extreme road conditions of a pit, even if the vehicle is triggered to decelerate, severe jolt and even damage of the vehicle cannot be avoided; the other is a self-vehicle control technology based on active safety road condition advanced perception; however, the detection performance of vision on the road surface fluctuation and the pits still needs to be improved, especially when the vehicle is driven at night in rainy days with poor sight, the vision perception is still difficult to deal with for the pits with full water, and meanwhile, the technology is limited by the performance of the sensor; in order to solve the problems, the invention provides a rough road condition identification and vehicle control method based on V2X and visual fusion.

Disclosure of Invention

The present invention is directed to solving at least one of the problems of the prior art. Therefore, the invention provides the rough road condition identification and vehicle control method based on the fusion of V2X and vision, which realizes the accurate identification of rough road conditions and can adjust the vehicle speed or adjust the driving path to reduce or avoid discomfort or vehicle damage caused by the rough road.

In order to achieve the above object, an embodiment according to the first aspect of the present invention provides a rough road condition identification and vehicle control method based on V2X and visual fusion, which includes five sets of signal links, and includes the following specific steps:

the method comprises the following steps: acquiring original pavement data of each signal link;

step two: screening pits or bulges which are larger than a threshold value aiming at road surface data obtained by detecting three groups of signal links of V2I, V2V and vision-based road surface fluctuation perception; if the target exists, the final international flatness index IRI is obtained_fAssigning a value of 10, and directly entering the step six; otherwise, entering the third step;

step three: classifying the road surface data in the step two according to the signal link, respectively processing the road surface data by adopting a direct detection method, and sequentially obtaining the international flatness index IRI according to the road surface data calculation on the extension lines of the left and right wheel centers and the whole vehicle center along the vehicle head direction₁、IRI₂And IRI₃；

Step four: processing front vehicle lamp vertical position jumping data acquired by front vehicle lamp vertical direction position perception based on vision by adopting a response type detection method to obtain an international flatness index IRI₄；

Step five: weighting the international flatness indexes obtained by different signal links to obtain the final international flatness index IRI_f；

Step six: combined with the final international flatness index IRI_fThe international flatness index evaluation index is used for optimizing and adjusting the current running speed or running track; the international flatness index evaluation index comprises a comparison table for dividing the international flatness index and the road surface condition.

Further, the five groups of signal links are respectively: the link I is V2I communication, the link II is V2V communication, the link III is visual-based road surface fluctuation perception, the link II is visual-based front vehicle lamp vertical direction position perception, and the link II is control signals which are processed by combining the self vehicle with front end perception information through fusion, decision, planning and control algorithms and are sent to corresponding execution units.

Further, the direct assay in step three is represented as: the international flatness index of the current road is obtained by calculating the undulation condition of the cross section of the road, and the international flatness index comprises the following specific steps:

dividing each lane into 3 interested areas along the driving direction of the vehicle, and applying a general statistical method to calculate the flatness of the current road, wherein the calculation formula is as follows:

wherein n is the number of elevation values contained in each step length; y is_iIs the ith elevation value (mm) in each step length; y is_aCalculating an average value of the elevation values contained in each step length;

s is a road flatness index expressed by a standard deviation, and is converted into an international flatness index IRI according to the following conversion formula_a：

Further, the 3 regions of interest include a region on a center line of the left wheel, a region on a center line of the right wheel, and a region on an extension line of the center of the entire vehicle in the direction of the vehicle head.

Further, the responsive detection method in step four is represented as: the international flatness index of the current road is deduced by detecting the mechanical response of the vehicle to the rough road surface, and the method specifically comprises the following steps:

where v is the absolute velocity of the vehicle relative to the ground, Δ t is the time interval of road sampling, l_iFor discrete samplingThe distance between the fixed point of the vehicle body and the ground is obtained, n is the number of sampling points, and the sampling distance of the road can be known by combining delta t.

Further, the final international flatness index IRI_fThe calculation formula of (a) is as follows:

wherein IRI₁，IRI₂，IRI₃，IRI₄，ω₁，ω₂，ω₃，ω₄The international flatness indexes and the corresponding weight values of the international flatness indexes are obtained by the four groups of signal links, namely V2I, V2V, road surface fluctuation perception based on vision and front vehicle lamp vertical direction position perception based on vision.

Further, when the signal link is not present, the weight value thereof takes 0.

Further, in the sixth step, the optimizing and adjusting of the current running speed or running track specifically includes:

when IRI_fWhen the value is less than 4, no limitation is carried out on the speed of the whole vehicle;

when IRI_fWhen the value is between 4 and 6, the speed is not limited, but text or voice reminding is carried out through man-machine interaction;

when IRI_fWhen the value is more than 6, limiting the speed of the whole vehicle by 30 km/h; and for the automatically driven vehicle, path re-planning is synchronously carried out, and when a new driving path is available, the path is updated and the whole vehicle is controlled to move up and down and avoid.

Compared with the prior art, the invention has the beneficial effects that:

according to the invention, the intelligent perception information of the single vehicle is fully mined by five groups of signal links, the real-time state of the road surface is perceived by using the forward-looking camera, the advantages and the disadvantages of each perception unit are comprehensively considered, the advantages of the V2X and the visual sensor are fully exerted by the fusion of the V2X and the vision, the defects of the perception units are overcome, and the fluctuating road condition is accurately and timely perceived and fed back, so that the extreme scenes which are difficult to be covered by the traditional single vehicle intelligence can be better covered, and the accuracy, the environmental adaptability and the robustness of the whole vehicle to the road flatness perception are improved;

for a vehicle target on the road, the road fluctuation condition of the current position of the front vehicle is deduced by observing the position jump of a front vehicle lamp in the vertical direction and combining the speed of the front vehicle, the driving speed and the track of the self vehicle are adjusted by integrating the visual perception results of the road end and the target end, the safety and the comfort of driving under the fluctuant road surface are effectively improved, and the damage of a larger pit or a larger bulge to the vehicle can be avoided or reduced on the extremely fluctuant road surface.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a schematic block diagram of the present invention.

Fig. 2 is a schematic structural diagram of five groups of links in the present invention.

Detailed Description

The technical solutions of the present invention will be described below clearly and completely in conjunction with the embodiments, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1 to 2, the rough road condition recognition and vehicle control method based on V2X and vision fusion includes five signal links, wherein the link (i) is V2I communication, the link (ii) is V2V communication, the link (iii) is vision-based road surface rough sensing, the link (iv) is vision-based front vehicle lamp vertical position sensing, and the link (iv) is a control signal obtained by performing fusion, decision, planning, control, etc. on the self vehicle combined with front-end sensing information and then sending the control signal to a corresponding execution unit; wherein, the evaluation standards of the road undulation condition are all unified to the international flatness index (IRI) for calculation; the method comprises the following specific steps:

the method comprises the following steps: acquiring original pavement data of each link;

step two: screening pits or bulges which are larger than a threshold value aiming at data obtained by V2I, V2V and vision-based road surface undulation perception detection; if there is a pit or a bump greater than the threshold, then IRI is applied_fAssigning a value of 10, and directly entering the step six;

step three: if no pit or bulge larger than the threshold exists, the rest road surface data is processed by adopting a direct detection method, and the international flatness index IRI is calculated by adopting the road surface data on the extension lines of the centers of the left wheel, the right wheel and the whole vehicle along the direction of the vehicle head₁、IRI₂And IRI₃(ii) a Wherein the direct detection method is represented by: a series of mathematical analyses are carried out by calculating the fluctuation condition of the cross section of the road to obtain the international flatness index of the current road, and the method specifically comprises the following steps:

dividing each lane into 3 interested areas along the driving direction of the vehicle, wherein the interested areas comprise an area on the central line of a left wheel, an area on the central line of a right wheel and an area on the extension line of the center of the whole vehicle along the direction of the head of the vehicle; the general statistical method is applied to calculate the flatness of the current road, and the calculation formula is as follows:

Wherein α ═ 1, 2, 3; refer to V2I, V2V and vision-based three ways of road surface undulation perception, respectively;

step four: processing front vehicle lamp vertical position jumping data acquired by front vehicle lamp vertical direction position perception based on vision by adopting a response type detection method, and calculating international flatness index IRI of a road surface₄(ii) a Wherein the responsive detection method is represented by: through the mechanics response of detecting the vehicle to the road surface that undulates, and then deduce the undulation condition of road, be international roughness index, specifically include:

calculating the undulation condition IRI of the current road surface by applying the following calculation formula_b：

Where v is the absolute velocity of the vehicle relative to the ground, Δ t is the time interval of road sampling, l_iObtaining the distance of the vehicle body fixing point vertical to the ground direction by discrete sampling, wherein n is the number of sampling points, and the sampling distance of the road can be known by combining delta t; wherein b takes the value 4;

in the embodiment, the link r is the vision-based position perception of the front vehicle lights in the vertical direction, is a big bright point of the invention, and realizes indirect perception of the road environment on the basis of fully exploiting the perception capability of the single vehicle. When the current target vehicle is on an undulating road surface, the position of the vehicle lamp in the vertical direction jumps, the undulating condition of the road surface where the vehicle lamp is located is deduced according to the second response type detection method, and therefore the driving strategy of the vehicle is adjusted within sufficient time. In addition, based on the signal, the defect of vision on pit depth detection can be effectively avoided, and more effective information is provided for a decision planning algorithm at the rear end;

step five: weighting all the obtained international flatness indexes to obtain the final international flatness index IRI_f(ii) a The specific calculation formula is as follows:

wherein IRI₁，IRI₂，IRI₃，IRI₄，ω₁，ω₂，ω₃，ω₄The international flatness indexes and the corresponding weight values are respectively obtained by four modes of V2I, V2V, vision-based road surface fluctuation perception and vision-based front vehicle lamp vertical direction position perception, and when the mode does not exist, the weight value is 0; in the present embodiment, for example: omega₁，ω₂，ω₃，ω₄Respectively taking 4, 2, 2 and 2;

step six: according to the final international flatness index, limiting the speed of the vehicle or re-planning the current driving path to avoid a severe bumpy road surface according to the performance of the vehicle in combination with the table 1, which specifically comprises the following steps:

table 1: international flatness index and pavement condition partitioning

When IRI_fWhen the value is less than 4, no limitation is made on the speed of the whole vehicle, wherein IRI_fUnit of (1) is m x km^-1(ii) a When IRI_fWhen the value is between 4 and 6, the speed is not limited, but text or voice reminding is carried out through human-machine interaction (HMI); when IRI_fWhen the value is more than 6, limiting the speed of the whole vehicle by 30 km/h; and for the automatically driven vehicle, path re-planning is synchronously carried out, and when a new driving path is available, the path is updated and the whole vehicle is controlled to move up and down and avoid.

According to the invention, through fully mining the intelligent sensing information of the single vehicle, the real-time state of the road surface is sensed by using the front-view camera, the advantages and the disadvantages of all sensing units are comprehensively considered, the advantages of the V2X and the visual sensor are fully exerted through the fusion of the V2X and the vision, the defects of the visual sensor are overcome, the fluctuating road condition is accurately and timely sensed and fed back, for the road vehicle target, the road fluctuating condition of the current position of the front vehicle is deduced through the observation of the position jump of the lamp of the front vehicle in the vertical direction and the combination of the speed of the front vehicle, the visual sensing results of the road end and the target end are synthesized, the running speed and the track of the self vehicle are adjusted, and the safety and the comfort of the vehicle running on the fluctuating road surface are effectively improved.

The working principle of the invention is as follows:

the rough road condition identification and vehicle control method based on V2X and vision fusion is characterized in that when in work, the original road surface data of each link are firstly obtained; screening pits or bulges which are larger than a threshold value aiming at data obtained by V2I, V2V and vision-based road surface fluctuation perception detection; if there is a pit or a bump greater than the threshold, then IRI is applied_fAssigning a value of 10; if no pit or bulge larger than the threshold exists, the rest road surface data is processed by adopting a direct detection method, and the international flatness index IRI is calculated by adopting the road surface data on the extension lines of the centers of the left wheel, the right wheel and the whole vehicle along the direction of the vehicle head₁、IRI₂And IRI₃(ii) a Processing front vehicle lamp vertical position jumping data acquired by front vehicle lamp vertical direction position perception based on vision by adopting a response type detection method, and calculating international flatness index IRI of a road surface₄(ii) a Then all the obtained international flatness indexes are weighted to obtain the final flatness index IRI_f(ii) a According to the final international flatness index IRI_fThe speed of the self-vehicle is limited or the current driving path is planned again in combination with the performance of the self-vehicle so as to avoid a violent bumpy road surface, and the safety and the comfort of driving under the bumpy road surface are effectively improved.

In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise forms disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. The fluctuating road condition identification and vehicle control method based on V2X and vision fusion comprises five groups of signal links and is characterized by comprising the following specific steps:

step two: screening pits or bulges which are larger than a threshold value aiming at road surface data obtained by detecting three groups of signal links of V2I, V2V and vision-based road surface fluctuation perception; if such a target exists, the final international flatness index IRI will be_fAssigning 10, and directly entering the step six; otherwise, entering the third step;

Step five: weighting the international flatness indexes acquired from different sources to obtain the final international flatness index IRI_f；

Step six: combined with the final international flatness index IRI_fThe international flatness index evaluation index is used for optimizing and adjusting the current running speed or running track; china of ChinaThe evaluation index of the international flatness index is a comparison table divided by the international flatness index and the road surface condition.

2. The method for rough road condition identification and vehicle control based on V2X and visual fusion of claim 1, wherein the five sets of signal links are respectively: the link I is V2I communication, the link II is V2V communication, the link III is visual-based road surface fluctuation perception, the link II is visual-based front vehicle lamp vertical direction position perception, and the link II is control signals which are processed by combining the self vehicle with front end perception information through fusion, decision, planning and control algorithms and are sent to corresponding execution units.

3. The method for rough road condition recognition and vehicle control based on V2X and visual fusion of claim 1, wherein the direct detection method in step three is represented as: the international flatness index of the current road is obtained by calculating the undulation condition of the cross section of the road, and the international flatness index comprises the following specific steps:

wherein n is the number of elevation values contained in each step length; y is_iThe ith elevation value in each step length; y is_aCalculating an average value of the elevation values contained in each step length;

4. The rough road condition recognition and vehicle control method based on V2X and visual fusion as claimed in claim 3, wherein the 3 regions of interest include a region on the center line of the left wheel, a region on the center line of the right wheel, and a region on the extension line of the center of the entire vehicle in the direction of the vehicle head.

5. The rough road condition recognition and vehicle control method based on V2X and visual fusion of claim 1, wherein the step four responsive detection method is shown as: the international flatness index of the current road is deduced by detecting the mechanical response of the vehicle to the rough road surface, and the method specifically comprises the following steps:

where v is the absolute velocity of the vehicle relative to the ground, Δ t is the time interval of road sampling, l_iThe distance of the vehicle body fixing point perpendicular to the ground direction is obtained through discrete sampling, n is the number of sampling points, and the sampling distance of the road can be known through combining delta t.

6. The rough road condition identification and vehicle control method based on V2X and visual fusion as claimed in claim 1, wherein the final international flatness index IRI_fThe calculation formula of (a) is as follows:

7. The method as claimed in claim 6, wherein the weighting value is 0 when the signal link is not present.

8. The rough road condition recognition and vehicle control method based on V2X and visual fusion as claimed in claim 1, wherein the step six of optimizing and adjusting the current driving speed or driving track specifically comprises: