阅读说明：本技术 智能驾驶车辆面对目标车辆切入的行车场景的识别方法 (Method for identifying driving scene cut into by intelligent driving vehicle facing target vehicle ) 是由 胡杰 张敏超 陈瑞楠 钟鑫凯 朱令磊 徐文才 颜伏伍 于 2021-08-04 设计创作，主要内容包括：本发明公开了一种智能驾驶车辆面对目标车辆切入的行车场景的识别方法,包括：将切入型典型行车场景进行分类、计算目标车辆预测轨迹与智能驾驶车辆未来行驶轨迹之间关系的特征值、基于特征值和分类的典型行车场景建立场景识别的逻辑判别关系、获取当前行驶的目标车辆的预测轨迹与当前行驶的智能驾驶车辆的未来行驶轨迹之间关系的特征值及将获取的特征值代入逻辑判别关系进行逻辑判别,以识别当前行驶的智能驾驶车辆的行车场景。本发明方法不仅能识别当前车辆的行车场景并能评估目标车辆切入智能驾驶车辆行驶轨迹的危险态势,而且计算量小、实时性好。(The invention discloses a method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle, which comprises the following steps: classifying cut-in type typical driving scenes, calculating a characteristic value of a relation between a predicted track of a target vehicle and a future driving track of the intelligent driving vehicle, establishing a logical discrimination relation of scene recognition based on the characteristic value and the classified typical driving scenes, obtaining the characteristic value of the relation between the predicted track of the current driving target vehicle and the future driving track of the current driving intelligent driving vehicle, and substituting the obtained characteristic value into the logical discrimination relation to perform logical discrimination so as to recognize the driving scene of the current driving intelligent driving vehicle. The method can identify the driving scene of the current vehicle and evaluate the dangerous situation of the target vehicle cutting into the driving track of the intelligent driving vehicle, and has small calculated amount and good real-time performance.)

1. A method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle is characterized by comprising the following steps:

switching the predicted track of the target vehicle into a typical driving scene of a future driving track of the intelligent driving vehicle for classification;

calculating a characteristic value of a relation between the predicted track of the target vehicle and the future travel track of the intelligent driving vehicle according to the predicted track information of the target vehicle and the future travel track information of the intelligent driving vehicle;

establishing a logical discrimination relation of scene recognition based on the characteristic values and the classified typical driving scenes;

acquiring a characteristic value of a relation between a predicted track of a currently running target vehicle and a future running track of a currently running intelligent driving vehicle;

and substituting the characteristic value of the relationship between the predicted track of the current running target vehicle and the future running track of the current running intelligent driving vehicle into the logic judgment relationship to carry out logic judgment, thereby identifying the driving scene of the current running intelligent driving vehicle.

2. The method according to claim 1, wherein the step of entering the predicted trajectory of the target vehicle into a typical driving scenario of a future trajectory of the smart driving vehicle for classification is embodied as:

the method comprises the following steps of (1) switching a predicted track of a target vehicle into a dangerous scene of a future driving track of an intelligent driving vehicle to be generalized into two dangerous scenes under a road coordinate system;

and based on the two dangerous scenes, switching the predicted track of the target vehicle into all typical driving scenes of the future driving track of the intelligent driving vehicle for classification.

3. The method of claim 2, wherein the two hazard scenarios are:

cutting into a dangerous scene of a future travel track of the intelligent driving vehicle from the front of the intelligent driving vehicle due to the predicted track of the target vehicle; and

the predicted trajectory of the target vehicle cuts into the dangerous scene of the future travel trajectory of the intelligent driving vehicle from the rear of the intelligent driving vehicle.

4. The method of claim 3, wherein:

the dangerous scene of the future driving track of the intelligent driving vehicle cut into by the target vehicle from the front of the intelligent driving vehicle comprises the following steps:

the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene;

the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out later to form a dangerous scene;

the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out;

the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the dangerous scene;

the dangerous scene cut into the future travel track of the intelligent driving vehicle from the rear of the intelligent driving vehicle due to the predicted track of the target vehicle comprises the following steps:

when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a dangerous scene of the future running track of the intelligent driving vehicle from the rear of the intelligent driving vehicle.

5. The method of claim 4, wherein all typical driving scenarios are classified as:

(1) a non-hazardous scene;

(2) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a non-dangerous scene;

(3) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene;

(4) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out of the non-dangerous scene;

(5) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out later to form a dangerous scene;

(6) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not cut out in the non-dangerous scene;

(7) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out;

(8) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the non-dangerous scene;

(9) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the dangerous scene;

(10) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a non-dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(11) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(12) the predicted trajectory of the target vehicle approaches quickly behind the smart driving vehicle and the smart driving vehicle is in a non-hazardous scene within the predicted trajectory of the target vehicle.

6. The method of claim 5, wherein the step of calculating the characteristic value of the relationship between the predicted trajectory of the target vehicle and the future travel trajectory of the smart driving vehicle based on the predicted trajectory information of the target vehicle and the future travel trajectory information of the smart driving vehicle specifically comprises:

obtaining the predicted track information of the target vehicle and the future driving track information of the intelligent driving vehicle through an intelligent driving vehicle prediction module;

and comparing and analyzing the time of reaching each point on the predicted track and the spatial coordinates of the point in the predicted track information of the target vehicle with the time of reaching each point on the future travel track and the spatial coordinates of the point in the future travel track information of the intelligent driving vehicle to obtain the characteristic value of the relationship between the predicted track of the target vehicle and the future travel track of the intelligent driving vehicle.

7. The method of claim 6, wherein the characteristic value of the relationship between the predicted trajectory of the target vehicle and the future travel trajectory of the smart driving vehicle comprises:

TTCR: on the predicted track of the target vehicle, the moment when the predicted track of the target vehicle cuts into the future travel track of the intelligent driving vehicle;

TTCRO: the time when the predicted track of the target vehicle is cut out of the future driving track of the intelligent driving vehicle on the predicted track of the target vehicle;

TTCI: the time when the predicted trajectory of the target vehicle cuts into the future travel trajectory of the intelligent driving vehicle on the future travel trajectory of the intelligent driving vehicle;

TTCO: the time when the predicted track of the target vehicle cuts out the future running track of the intelligent driving vehicle on the future running track of the intelligent driving vehicle;

TTC: and intelligently driving the time when the vehicle collides with the target vehicle.

8. The method of claim 7, wherein the logical discriminant relationship of scene recognition is specifically:

(1) non-dangerous scene: TTCR ═ -1, TTCRO ═ 1, TTCI ═ 1, TTCO ═ 1;

(2) a non-dangerous scene that the predicted track of the target vehicle is within the future driving track of the intelligent driving vehicle and is not cut out: TTCR >0, TTCI ═ 1, TTC <0, with the proviso TTCRO ═ -1| | | TTCO ═ 1;

(3) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene: TTCR >0, TTCI ═ 1, TTC >0& & TTC < X, additional condition TTCRO ═ -1| | | TTCO ═ 1;

(4) a non-dangerous scene in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart driving vehicle and is subsequently cut out: TTCR >0, TTCRO >0, TTCI ═ 1, TTCO >0, additional condition (TTC <0) | (TTC >0& & TTCO < TTCRO);

(5) dangerous scenes in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart driving vehicle and is cut out later: TTCR >0, TTCRO >0, TTCI ═ 1, TTCO >0, TTC >0& & TTC < X, additional condition TTCO > ═ TTCRO;

(6) a non-dangerous scene in which the predicted trajectory of the target vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle and is not cut out: TTCR >0, TTCRO ═ 1, TTCI >0, TTCO ═ 1, TTC < 0;

(7) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out: TTCR >0, TTCRO ═ 1, TTCI >0, TTCO ═ 1, TTC >0& & TTC < X;

(8) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is switched out of the non-dangerous scene: TTCR >0, TTCRO >0, TTCI >0, TTCO >0, and TTCRO > -TTCO;

(9) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is switched out of the dangerous scene: TTCR >0, TTCRO >0, TTCI >0, TTCO >0, TTC >0& & TTC < X, additional condition TTCRO < TTCO;

(10) when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a non-dangerous scene of the future running track of the intelligent driving vehicle from the rear of the intelligent driving vehicle: TTCR >0, TTCI >0, TTC < 0;

(11) when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle, wherein TTCR is >0, TTCI is >0, TTC is >0& & TTC < X;

(12) a non-dangerous scene in which the predicted trajectory of the target vehicle approaches quickly behind the intelligent driving vehicle and the intelligent driving vehicle is in the predicted trajectory of the target vehicle, wherein TTCR is-1 and TTCI is > 0;

where "-1" indicates that this characteristic value is not present, and X is the threshold value for TTC.

9. The method according to claim 1, wherein the step of obtaining the characteristic value of the relationship between the predicted trajectory of the currently traveling target vehicle and the future traveling trajectory of the currently traveling smart driving vehicle is embodied as:

obtaining the predicted track information of the current running target vehicle and the future running track information of the current running intelligent driving vehicle through an intelligent driving vehicle prediction module;

and comparing and analyzing the time of reaching each point on the predicted track in the predicted track information of the current running target vehicle and the spatial coordinates of the point with the time of reaching each point on the future running track and the spatial coordinates of the point in the future running track information of the current running intelligent driving vehicle to obtain the characteristic value of the relationship between the predicted track of the current running target vehicle and the future running track of the current running intelligent driving vehicle.

Technical Field

The invention relates to the field of decision planning of intelligent driving systems, in particular to a method for identifying a driving scene cut by an intelligent driving vehicle facing a target vehicle.

Background

The intelligent driving vehicle intensively applies the technologies of computer, modern sensing, information fusion, communication, artificial intelligence, automatic control and the like, and is a typical high and new technology complex. In recent years, intelligent vehicles have become hot spots for the research in the field of vehicle engineering in the world and new power for the growth of the automobile industry, and many developed countries incorporate the intelligent vehicles into intelligent transportation systems which are intensively developed.

The key common technology of the intelligent driving vehicle comprises three parts of environment perception, decision planning and control execution. In particular, the context awareness module is responsible for detecting the environment surrounding the vehicle, such as the vehicle, the pedestrian, the road sign, etc. And the decision planning module is responsible for identifying and analyzing the information detected by the environment sensing module, making a decision and issuing a next action instruction. The control execution layer is responsible for controlling the execution mechanisms such as an accelerator pedal and a steering wheel of the vehicle to output and run according to the expectation. In the decision planning module, scene recognition technology is a very important part, and has an important influence on the safety of the vehicle.

At present, a scene recognition method of an intelligent driving vehicle mainly utilizes massive data to train a deep learning or machine learning model for recognition. However, this method requires collection of a large amount of scene data, and at the same time, has a very high demand on the computational power of the on-board controller, and thus it is difficult to practically apply it to an intelligent driving vehicle.

Disclosure of Invention

The invention provides a method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle, which can identify the driving scene where the current vehicle is located and better evaluate the dangerous situation of the driving track cut into the intelligent driving vehicle by the target vehicle, and has the advantages of small calculated amount and good real-time property.

In order to achieve the purpose, the invention provides a method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle, which comprises the following steps:

switching the predicted track of the target vehicle into a typical driving scene of a future driving track of the intelligent driving vehicle for classification;

calculating a characteristic value of a relation between the predicted track of the target vehicle and the future travel track of the intelligent driving vehicle according to the predicted track information of the target vehicle and the future travel track information of the intelligent driving vehicle;

establishing a logical discrimination relation of scene recognition based on the characteristic values and the classified typical driving scenes;

acquiring a characteristic value of a relation between a predicted track of a currently running target vehicle and a future running track of a currently running intelligent driving vehicle;

and substituting the characteristic value of the relationship between the predicted track of the current running target vehicle and the future running track of the current running intelligent driving vehicle into the logic judgment relationship to carry out logic judgment, thereby identifying the driving scene of the current running intelligent driving vehicle.

Preferably, the step of classifying the predicted trajectory of the target vehicle into a typical driving scene of a future driving trajectory of the smart driving vehicle is specifically:

the method comprises the following steps of (1) switching a predicted track of a target vehicle into a dangerous scene of a future driving track of an intelligent driving vehicle to be generalized into two dangerous scenes under a road coordinate system;

and based on the two dangerous scenes, switching the predicted track of the target vehicle into all typical driving scenes of the future driving track of the intelligent driving vehicle for classification.

Preferably, the two danger scenarios are:

cutting into a dangerous scene of a future travel track of the intelligent driving vehicle from the front of the intelligent driving vehicle due to the predicted track of the target vehicle; and

the predicted trajectory of the target vehicle cuts into the dangerous scene of the future travel trajectory of the intelligent driving vehicle from the rear of the intelligent driving vehicle.

Preferably, the dangerous scene of the future travel track of the intelligent driving vehicle cut into by the target vehicle from the front of the intelligent driving vehicle comprises:

the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene;

the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out later to form a dangerous scene;

the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out;

the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the dangerous scene;

the dangerous scene cut into the future travel track of the intelligent driving vehicle from the rear of the intelligent driving vehicle due to the predicted track of the target vehicle comprises the following steps:

when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a dangerous scene of the future running track of the intelligent driving vehicle from the rear of the intelligent driving vehicle.

Preferably, all typical driving scenarios are classified as:

(1) a non-hazardous scene;

(2) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a non-dangerous scene;

(3) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene;

(4) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out of the non-dangerous scene;

(5) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out later to form a dangerous scene;

(6) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not cut out in the non-dangerous scene;

(7) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out;

(8) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the non-dangerous scene;

(9) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the dangerous scene;

(10) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a non-dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(11) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(12) the predicted trajectory of the target vehicle approaches quickly behind the smart driving vehicle and the smart driving vehicle is in a non-hazardous scene within the predicted trajectory of the target vehicle.

Preferably, the step of calculating the characteristic value of the relationship between the predicted trajectory of the target vehicle and the future travel trajectory of the smart driving vehicle according to the predicted trajectory information of the target vehicle and the future travel trajectory information of the smart driving vehicle specifically includes:

obtaining the predicted track information of the target vehicle and the future driving track information of the intelligent driving vehicle through an intelligent driving vehicle prediction module;

and comparing and analyzing the time of reaching each point on the predicted track and the spatial coordinates of the point in the predicted track information of the target vehicle with the time of reaching each point on the future travel track and the spatial coordinates of the point in the future travel track information of the intelligent driving vehicle to obtain the characteristic value of the relationship between the predicted track of the target vehicle and the future travel track of the intelligent driving vehicle.

Preferably, the characteristic value of the relationship between the predicted trajectory of the target vehicle and the future travel trajectory of the smart driving vehicle includes:

TTCR: on the predicted track of the target vehicle, the moment when the predicted track of the target vehicle cuts into the future travel track of the intelligent driving vehicle;

TTCRO: the time when the predicted track of the target vehicle is cut out of the future driving track of the intelligent driving vehicle on the predicted track of the target vehicle;

TTCI: the time when the predicted trajectory of the target vehicle cuts into the future travel trajectory of the intelligent driving vehicle on the future travel trajectory of the intelligent driving vehicle;

TTCO: the time when the predicted track of the target vehicle cuts out the future running track of the intelligent driving vehicle on the future running track of the intelligent driving vehicle;

TTC: and intelligently driving the time when the vehicle collides with the target vehicle.

Preferably, the logical discrimination relationship of scene recognition is specifically:

(1) non-dangerous scene: TTCR ═ -1, TTCRO ═ 1, TTCI ═ 1, TTCO ═ 1;

(2) a non-dangerous scene that the predicted track of the target vehicle is within the future driving track of the intelligent driving vehicle and is not cut out: TTCR >0, TTCI ═ 1, TTC <0, with the proviso TTCRO ═ -1| | | TTCO ═ 1;

(3) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene: TTCR >0, TTCI ═ 1, TTC >0& & TTC < X, additional condition TTCRO ═ -1| | | TTCO ═ 1;

(4) a non-dangerous scene in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart driving vehicle and is subsequently cut out: TTCR >0, TTCRO >0, TTCI ═ 1, TTCO >0, additional condition (TTC <0) | (TTC >0& & TTCO < TTCRO);

(5) dangerous scenes in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart driving vehicle and is cut out later: TTCR >0, TTCRO >0, TTCI ═ 1, TTCO >0, TTC >0& & TTC < X, additional condition TTCO > ═ TTCRO;

(6) a non-dangerous scene in which the predicted trajectory of the target vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle and is not cut out: TTCR >0, TTCRO ═ 1, TTCI >0, TTCO ═ 1, TTC < 0;

(7) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out: TTCR >0, TTCRO ═ 1, TTCI >0, TTCO ═ 1, TTC >0& & TTC < X;

(8) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is switched out of the non-dangerous scene: TTCR >0, TTCRO >0, TTCI >0, TTCO >0, and TTCRO > -TTCO;

(9) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is switched out of the dangerous scene: TTCR >0, TTCRO >0, TTCI >0, TTCO >0, TTC >0& & TTC < X, additional condition TTCRO < TTCO;

(10) when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a non-dangerous scene of the future running track of the intelligent driving vehicle from the rear of the intelligent driving vehicle: TTCR >0, TTCI >0, TTC < 0;

(11) when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into a dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle, wherein TTCR is >0, TTCI is >0, TTC is >0& & TTC < X;

(12) a non-dangerous scene in which the predicted trajectory of the target vehicle approaches quickly behind the intelligent driving vehicle and the intelligent driving vehicle is in the predicted trajectory of the target vehicle, wherein TTCR is-1 and TTCI is > 0;

where "-1" indicates that this characteristic value is not present, and X is the threshold value for TTC.

Preferably, the step of obtaining the characteristic value of the relationship between the predicted trajectory of the currently running target vehicle and the future running trajectory of the currently running intelligent driving vehicle specifically includes:

obtaining the predicted track information of the current running target vehicle and the future running track information of the current running intelligent driving vehicle through an intelligent driving vehicle prediction module;

and comparing and analyzing the time of reaching each point on the predicted track in the predicted track information of the current running target vehicle and the spatial coordinates of the point with the time of reaching each point on the future running track and the spatial coordinates of the point in the future running track information of the current running intelligent driving vehicle to obtain the characteristic value of the relationship between the predicted track of the current running target vehicle and the future running track of the current running intelligent driving vehicle.

The technical scheme provided by the invention has the beneficial effects that:

the method establishes a logical judgment relation of scene recognition based on the characteristic value of the relation between the predicted track of the target vehicle and the future driving track of the intelligent driving vehicle and the classified typical driving scene, and can recognize the driving scene of the intelligent driving vehicle in real time only by substituting the characteristic value of the relation between the predicted track of the target vehicle and the future driving track of the intelligent driving vehicle into the logical judgment relation for logical judgment when the scene of the intelligent driving vehicle is recognized.

Drawings

Fig. 1 is a flowchart of a method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle according to an embodiment of the present invention;

2A-2D illustrate four of the hazardous scenarios for a target vehicle cutting into the smart-driven vehicle's future travel trajectory from ahead of the smart-driven vehicle;

FIG. 3 illustrates one of the hazardous scenarios for a predicted trajectory of a target vehicle cutting into a future travel trajectory of a smart-driven vehicle from behind the smart-driven vehicle.

Detailed Description

In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be further described in detail below with reference to the drawings in the embodiments of the present invention.

As shown in fig. 1, an embodiment of the present invention provides a method for identifying a driving scene cut into by an intelligent driving vehicle facing a target vehicle, including the following steps:

step S1: the method comprises the following steps of (1) closing a predicted track of a target vehicle into a dangerous scene of a future driving track of an intelligent driving vehicle to be generalized into two dangerous scenes under a road coordinate system, wherein the two dangerous scenes are as follows: (a) a danger scenario that cuts into the smart-driven vehicle's future travel trajectory from the front of the smart-driven vehicle due to the target vehicle's predicted trajectory and (b) a danger scenario that cuts into the smart-driven vehicle's future travel trajectory from the rear of the smart-driven vehicle due to the target vehicle's predicted trajectory;

specifically, (a) the risk scenario in which the predicted trajectory of the target vehicle cuts into the future travel trajectory of the smart-driven vehicle from the front of the smart-driven vehicle includes: (i) a dangerous scene in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart-driving vehicle and is not cut out (i.e., the predicted trajectory of the target vehicle is within the future travel trajectory of the smart-driving vehicle while remaining within the future travel trajectory of the smart-driving vehicle); (ii) a dangerous scene in which the predicted trajectory of the target vehicle is within the future travel trajectory of the smart driving vehicle and is cut out later (i.e., the predicted trajectory of the target vehicle is within the predicted trajectory of the smart driving vehicle while the target vehicle is ready to change lanes or accelerate, at which time the predicted trajectory of the target vehicle is cut out of the future travel trajectory of the smart driving vehicle); (iii) a dangerous scene in which the predicted trajectory of the target vehicle is not cut out after the predicted trajectory of the smart driving vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle (i.e., the predicted trajectory of the target vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle and then remains within the future travel trajectory of the smart driving vehicle); (iv) and a dangerous scene in which the predicted trajectory of the target vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle and then cut out (the predicted trajectory of the target vehicle is cut into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle and then the future travel trajectory of the smart driving vehicle is cut out). (b) The dangerous scene cut into the future travel track of the intelligent driving vehicle from the rear of the intelligent driving vehicle due to the predicted track of the target vehicle comprises the following steps: when the intelligent driving vehicle changes lanes, the predicted track of the target vehicle is switched into the dangerous scene of the future travel track of the intelligent driving vehicle from the rear of the intelligent driving vehicle (namely, when the intelligent driving vehicle changes lanes, the speed of the rear target vehicle is higher, and the predicted track of the target vehicle is switched into the future travel track of the intelligent driving vehicle from the rear of the intelligent driving vehicle).

In order to more intuitively understand the divided dangerous scenes, fig. 2A to 2D illustrate four dangerous scenes, i.e., the dangerous scene (i) described above, the dangerous scene (ii) described above, the dangerous scene (iii) described above, and the dangerous scene (iv) described above, respectively, in which the target vehicle cuts into the future travel trajectory of the smart driving vehicle from the front of the smart driving vehicle. FIG. 3 illustrates, for example, a dangerous scenario in which a predicted trajectory of a target vehicle cuts into a future travel trajectory of a smart-driven vehicle from behind the smart-driven vehicle. "t 0" in fig. 2A to 2D and fig. 3 is a future travel locus of the smart driving vehicle, and "t 1" is a predicted locus of the target vehicle.

Step S2: based on the two dangerous scenes, the predicted trajectory of the target vehicle is switched into all typical driving scenes of the future driving trajectory of the intelligent driving vehicle for classification, and all the typical driving scenes are classified into 12 types, including:

(1) a non-hazardous scene;

(2) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a non-dangerous scene;

(3) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is not cut out in a dangerous scene;

(4) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out of the non-dangerous scene;

(5) the predicted track of the target vehicle is in the future driving track of the intelligent driving vehicle and is cut out later to form a dangerous scene;

(6) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not cut out in the non-dangerous scene;

(7) the predicted track of the target vehicle is switched into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then is not switched out;

(8) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the non-dangerous scene;

(9) the predicted track of the target vehicle is cut into the future running track of the intelligent driving vehicle from the front of the intelligent driving vehicle and then cut out of the dangerous scene;

(10) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a non-dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(11) when the intelligent driving vehicle changes lanes, switching the predicted track of the target vehicle into a dangerous scene of a future driving track of the intelligent driving vehicle from the rear of the intelligent driving vehicle;

(12) the predicted trajectory of the target vehicle approaches quickly behind the smart driving vehicle and the smart driving vehicle is in a non-hazardous scene within the predicted trajectory of the target vehicle.

Step S3: and obtaining the predicted track information of the target vehicle and the future driving track information of the intelligent driving vehicle through an intelligent driving vehicle prediction module.

It should be noted that the predicted trajectory of the target vehicle and the future travel trajectory of the intelligent driving vehicle are output in the form of a spatio-temporal trajectory, and each point on the spatio-temporal trajectory, i.e. the trajectory, includes the time when the vehicle reaches the point and the spatial coordinates (i.e. the position) of the point in the coordinate system. The intelligent driving vehicle prediction module is a conventional intelligent driving vehicle prediction module, which is not described in detail herein.

Step S4: comparing and analyzing the time when the vehicle reaches each point on the track and the space coordinates of each point in the coordinate system in the predicted track information of the target vehicle with the time when the vehicle reaches each point on the track and the space coordinates of each point in the coordinate system in the future travel track information of the intelligent driving vehicle to obtain the characteristic value of the relation between the predicted track of the target vehicle and the future travel track of the intelligent driving vehicle;

specifically, the characteristic values include:

ttcr (time to cross): on the predicted track of the target vehicle, the moment when the predicted track of the target vehicle cuts into the future travel track of the intelligent driving vehicle;

TTCRO (Time to cross-off): the time when the predicted track of the target vehicle is cut out of the future driving track of the intelligent driving vehicle on the predicted track of the target vehicle;

TTCI (Time to cut-in): the time when the predicted trajectory of the target vehicle cuts into the future travel trajectory of the intelligent driving vehicle on the future travel trajectory of the intelligent driving vehicle;

TTCO (Time to cut-out): the time when the predicted track of the target vehicle cuts out the future running track of the intelligent driving vehicle on the future running track of the intelligent driving vehicle;

ttc (time to fusion): and intelligently driving the time when the vehicle collides with the target vehicle.

For clear understanding of the meaning of the above characteristic values, TTCI is described here as an example. TTCI indicates that the predicted trajectory of the target vehicle cuts into the future travel trajectory of the smart-driven vehicle at a time on the future travel trajectory of the smart-driven vehicle, that is, whether the predicted trajectory of the target vehicle coincides with the position of the smart-driven vehicle at a time point in the future is searched for on the future travel trajectory of the smart-driven vehicle, and if so, the predicted trajectory of the target vehicle is considered to cut into the future travel trajectory of the smart-driven vehicle, and the time is denoted as TTCI.

Step S5: based on the characteristic values TTCR, TTCRO, TTCI, TTCO, TTC and the classified 12 types of typical driving scenes, a logical discriminant relationship of scene recognition is established, wherein the established logical discriminant relationship is shown in the following table 1:

TABLE 1 logical discriminant relationship table

In table 1, "-1" indicates that the characteristic value does not exist, "\" indicates that the characteristic value does not need to be considered, "X" is a threshold value of TTC (the threshold value is determined by actual circumstances), "═ corresponds to the number," & "corresponds to the sum, and" additional condition "indicates that whether an additional condition is satisfied or not is judged again under the condition that the previous parameters are satisfied.

Step S6: and obtaining the predicted track information of the current running target vehicle and the future running track information of the current running intelligent driving vehicle through an intelligent driving vehicle prediction module.

Step S7: and comparing and analyzing the predicted track information of the currently running target vehicle and the future running track information of the currently running intelligent driving vehicle to obtain a characteristic value of the relation between the predicted track of the currently running target vehicle and the future running track of the currently running intelligent driving vehicle, wherein the predicted track information of the currently running target vehicle and the future running track information of the currently running intelligent driving vehicle comprise the time when the vehicle reaches each point on the track and the space coordinates of each point in a coordinate system.

Step S8: and substituting the characteristic value of the relation between the predicted track of the currently running target vehicle and the future running track of the currently running intelligent driving vehicle into the logic discrimination table established in the step S5, thereby identifying the driving scene of the currently running intelligent driving vehicle.

It can be understood that, in the embodiment, the process of predicting the motion trajectory of the target vehicle is to use the motion response information of the target vehicle after smoothing by the filter algorithm such as kalman filtering and the like as the input, predict the motion trajectory required by the target vehicle in the next N seconds by using a linear prediction method, draw the motion trajectory on the environment map and give a time attribute to the motion trajectory, so that the space-time motion trajectory (i.e., the predicted trajectory) of the vehicle in a period of time in the future can be obtained.

The present invention is not limited to the above preferred embodiments, and any modifications, equivalent replacements, improvements, etc. within the spirit and principle of the present invention should be included in the protection scope of the present invention.