CN107945574B

CN107945574B - Vehicle collision early warning method, device and equipment

Info

Publication number: CN107945574B
Application number: CN201711015264.4A
Authority: CN
Inventors: 邰冲
Original assignee: Neusoft Corp
Current assignee: Neusoft Corp
Priority date: 2017-10-25
Filing date: 2017-10-25
Publication date: 2020-04-10
Anticipated expiration: 2037-10-25
Also published as: CN107945574A

Abstract

The embodiment of the application discloses a vehicle collision early warning method, which comprises the steps of obtaining a real-time driving state of a target vehicle and a mapping relation between a historical driving state and a prior collision threat coefficient, then obtaining the prior collision threat coefficient of the target vehicle according to the real-time driving state of the target vehicle and the mapping relation, then calculating a posterior collision threat coefficient of the target vehicle according to the real-time driving state of the target vehicle, and finally judging whether early warning is needed to be carried out on the target vehicle according to the prior collision threat coefficient and the posterior collision threat coefficient. The prior collision coefficient can reflect the possibility of collision of the training vehicle in the historical driving state, so the combination of the prior collision threat coefficient and the posterior collision threat coefficient can more accurately predict the possibility of collision of the target vehicle with other surrounding vehicles, and the early warning error rate of the target vehicle is reduced.

Description

Vehicle collision early warning method, device and equipment

Technical Field

The application relates to the field of intelligent vehicles, in particular to a vehicle collision early warning method, device and equipment.

Background

In order to reduce the occurrence of traffic accidents, the prior art estimates the time of collision between vehicles based on real-time driving state parameters such as the speed, the driving direction, the relative distance between vehicles and the like of the vehicles, and if the estimated time of collision is less than or equal to a certain threshold, an early warning is performed to remind the driver to take corresponding anti-collision measures.

However, the early warning effect of the collision early warning method based on the real-time driving state of the vehicle in the prior art is not good, and the driving state of the vehicle at a certain moment cannot represent the driving trend of the vehicle, so that early warning errors may be caused. For example, if the instantaneous speed of the vehicle at a certain moment is large, so that the calculated collision time is smaller than a threshold value, warning is performed at this moment; and the speed is smaller in the next period of time, so that the calculated collision time is larger than the threshold value, and then the early warning is not carried out. Therefore, the early warning caused by the large instantaneous speed is the early warning error.

Disclosure of Invention

In order to solve the problem of inaccurate early warning in the prior art, the application provides a vehicle collision early warning method, device and equipment to realize more accurate prediction of the possibility of collision between a target vehicle and other surrounding vehicles, so that the early warning error rate of the target vehicle is reduced

In a first aspect, the present application provides a vehicle collision warning method, including:

acquiring a real-time running state of a target vehicle and a mapping relation between a historical running state and a prior collision threat coefficient, wherein the prior collision threat coefficient reflects collision possibility of a plurality of training vehicles in the historical running state;

obtaining a prior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle and the mapping relation;

calculating a posterior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle, wherein the posterior collision threat coefficient reflects the possibility of collision between the target vehicle and other vehicles around the target vehicle;

and early warning is carried out on the target vehicle according to the prior collision threat coefficient and the posterior collision threat coefficient.

Optionally, the obtaining of the mapping relationship between the historical driving state and the prior collision threat coefficient includes:

acquiring historical driving states and historical collision results of the training vehicles;

and obtaining a prior collision threat coefficient in the historical driving state according to the historical driving states and the historical collision results of the training vehicles, so as to obtain a mapping relation between the historical driving states and the prior collision threat coefficient.

Optionally, the plurality of training vehicles includes a first training vehicle and a second training vehicle;

the historical travel state includes a relative speed, a relative distance, and a relative travel direction between the first training vehicle and the second training vehicle under a historical travel environment.

Optionally, the real-time driving state of the target vehicle includes a relative speed, a relative distance, and a relative driving direction between the target vehicle and other vehicles around the target vehicle in the current driving environment.

Optionally, the historical driving environment and the current driving environment respectively include at least one of the following:

weather conditions, road shape, and road surface conditions.

Optionally, the road shape comprises: straight, turning, or crossing.

Optionally, the historical collision result includes one of the following:

crash and no airbag ejection, crash and airbag ejection, no crash and emergency braking, no crash and deceleration to 0 by braking and no crash and no deceleration to 0 by braking.

Optionally, the real-time driving state of the target vehicle includes real-time position information;

the calculating the posterior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle comprises:

predicting collision time of the target vehicle and other vehicles around the target vehicle according to the real-time position information of the target vehicle and the real-time position information of other vehicles around the target vehicle;

and obtaining the posterior collision threat coefficient according to the collision time.

Optionally, the predicting the collision time of the target vehicle and other vehicles around the target vehicle according to the real-time position information of the target vehicle and the real-time position information of other vehicles around the target vehicle includes:

predicting the position of the target vehicle at a future preset time according to the real-time position information of the target vehicle, and determining a safety area of the target vehicle based on the position of the target vehicle at the future preset time;

predicting the positions of the other vehicles arriving at the future preset time according to the real-time position information of the other vehicles around, and determining safety regions of the other vehicles around based on the positions of the other vehicles arriving at the future preset time;

and if the safety zone of the target vehicle is overlapped with the safety zones of other vehicles around, determining the time period from the current time to the future preset time as the collision time.

Optionally, the real-time driving state of the target vehicle further includes: the speed information, the yaw information and the current driving environment information of the target vehicle;

the predicting the position where the target vehicle arrives at the future preset time according to the real-time position information of the target vehicle comprises:

and predicting the position of the target vehicle at a future preset moment according to the real-time position information, the speed information, the yaw information and the current running environment information of the target vehicle.

Optionally, the speed information includes: the current speed and the current acceleration, the yaw information comprises a yaw angle, and the current driving environment information comprises a ground friction coefficient;

the predicting the position of the target vehicle at a future preset time according to the real-time position information, the speed information, the yaw information and the current running environment information of the target vehicle comprises:

and if the target vehicle moves straight, predicting the position of the target vehicle at a future preset moment according to the current speed, the current acceleration, the yaw angle and the ground friction coefficient.

Optionally, the speed information includes: the current speed and the current acceleration, the yaw information comprises a yaw angle and a yaw angular speed, and the current running environment information comprises a ground friction coefficient;

and if the target vehicle turns, predicting the position of the target vehicle at a future preset moment according to the current speed, the current acceleration, the yaw angle, the yaw angular speed and the ground friction coefficient.

In a second aspect, the present application provides a vehicle collision warning apparatus, the apparatus comprising:

the system comprises a first acquisition unit, a second acquisition unit and a third acquisition unit, wherein the first acquisition unit is used for acquiring a real-time running state of a target vehicle and a mapping relation between a historical running state and a prior collision threat coefficient, and the prior collision threat coefficient reflects the collision possibility of a plurality of training vehicles in the historical running state;

the second obtaining unit is used for obtaining a prior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle and the mapping relation;

the computing unit is used for computing a posterior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle, wherein the posterior collision threat coefficient reflects the possibility of collision between the target vehicle and other vehicles around the target vehicle;

and the early warning unit is used for early warning the target vehicle according to the prior collision threat coefficient and the posterior collision threat coefficient.

Optionally, the first obtaining unit includes:

the first acquisition subunit is used for acquiring historical driving states and historical collision results of the training vehicles;

and the second obtaining subunit is used for obtaining a priori collision threat coefficient in the historical driving state according to the historical driving states of the training vehicles and the historical collision results, so as to obtain a mapping relation between the historical driving states and the priori collision threat coefficient.

weather conditions, road shape, and road surface conditions.

Optionally, the road shape comprises: straight, turning, or crossing.

Optionally, the historical collision result includes one of the following:

the calculation unit includes:

a prediction subunit configured to predict a collision time between the target vehicle and other vehicles around the target vehicle, based on the real-time position information of the target vehicle and the real-time position information of the other vehicles around the target vehicle;

and the third obtaining subunit is used for obtaining the posterior collision threat coefficient according to the collision time.

Optionally, the predictor unit includes:

the first prediction module is used for predicting the position of the target vehicle at the future preset moment according to the real-time position information of the target vehicle;

a first determination module for determining a safety region of the target vehicle based on the position at which the target vehicle arrives at the future preset time predicted by the first prediction module;

the second prediction module is used for predicting the arrival positions of other surrounding vehicles at the future preset moment according to the real-time position information of the other surrounding vehicles;

a second determination module configured to determine a safety region of the other vehicle in the vicinity based on the position at which the other vehicle in the vicinity arrives at the future preset time predicted by the second prediction module;

and the third determining module is used for determining the time period from the current time to the future preset time as the collision time if the safety area of the target vehicle is overlapped with the safety areas of other vehicles around the target vehicle.

the first prediction module is specifically configured to predict the position where the target vehicle arrives at a future preset time according to the real-time position information, the speed information, the yaw information, and the current driving environment information of the target vehicle.

the first prediction module comprises:

and the first prediction submodule is used for predicting the position of the target vehicle at the future preset moment according to the current speed, the current acceleration, the yaw angle and the ground friction coefficient if the target vehicle moves straight.

the first prediction module comprises:

and the second prediction submodule is used for predicting the position of the target vehicle at the future preset moment according to the current speed, the current acceleration, the yaw angle, the yaw angular speed and the ground friction coefficient if the target vehicle turns.

In a third aspect, the present application provides a vehicle collision warning apparatus, the apparatus comprising:

a processor and a memory storing a program;

wherein the processor, when executing the program, performs the following:

According to the embodiment of the application, the real-time running state of the target vehicle and the mapping relation between the historical running state and the prior collision threat coefficient are obtained, then the prior collision threat coefficient of the target vehicle is obtained according to the real-time running state of the target vehicle and the mapping relation, then the posterior collision threat coefficient of the target vehicle is calculated according to the real-time running state of the target vehicle, and finally whether early warning needs to be carried out on the target vehicle is judged according to the prior collision threat coefficient and the posterior collision threat coefficient. Therefore, the method and the device have the advantages that the posterior threat collision threat coefficients which are obtained according to the real-time running state and reflect the possibility of collision between the target vehicle and other surrounding vehicles are referred, the prior collision threat coefficients which correspond to the historical running state which is the same as or similar to the real-time running state are referred, and the possibility of collision between the training vehicle and other surrounding vehicles can be more accurately predicted by combining the prior collision threat coefficients and the posterior collision threat coefficients, so that the early warning error rate of the target vehicle is reduced.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be 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 described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.

Fig. 1 is a schematic architecture diagram of a hardware scenario provided in the present application;

FIG. 2 is a schematic flow chart of a vehicle collision warning method provided by the present application;

FIG. 3 is a schematic diagram illustrating the intersection of two line segments provided herein;

fig. 4 is a block diagram of a vehicle collision warning device provided in the present application.

Detailed Description

In order to make the technical solutions of the present application better understood, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, 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 application.

The present application scheme is described below with reference to specific application scenarios, for example, one of the scenarios in the embodiment of the present application may be applied to a hardware scenario shown in fig. 1, where the hardware includes: a Road Side Unit (RSU) 101, an On Board Unit (OBU) 102, a Controller Area Network (CAN) 103, and a Global Positioning System (GPS) 104.

The roadside unit 101 may be configured to acquire a driving state of a passing vehicle, for example, the roadside unit 101 may acquire vehicle information of the passing vehicle by using a Dedicated Short Range Communications (DSRC), or may directly acquire vehicle information of the passing vehicle by performing communication connection with the on-board unit 102, where the vehicle information may include vehicle safety information and a driving state of the vehicle. After the vehicle information of the passing vehicle is acquired, the roadside unit 101 may further obtain a prior collision threat coefficient and a posterior collision threat coefficient of the passing vehicle by using the vehicle information, and determine whether to send alarm information to the on-board unit 102 according to the prior collision threat coefficient and the posterior collision threat coefficient.

The controller area network bus module 103 may be configured to acquire vehicle safety information of the vehicle, such as whether the vehicle is braked, whether emergency braking or an airbag is activated, and transmit the vehicle safety information to the on-board unit 102.

The gps module 104 may be configured to collect a vehicle driving status, such as a speed, a real-time location, and a driving direction of the vehicle, and transmit the vehicle driving status to the on-board unit 102.

The vehicle-mounted unit 102 CAN be used for receiving the vehicle safety information sent by the CAN bus module 103 and the vehicle running state sent by the GPS module 104 and sending the information to the roadside unit 101; and receiving alarm information sent by the roadside unit 101, and giving an alarm to the vehicle according to the alarm information.

The roadside unit 101 and the vehicle-mounted unit 102 can establish communication connection through DSRC technology.

In one implementation, the functions of the roadside unit 101 may be implemented by the on-board unit 102. It should be understood that the foregoing application scenarios are only shown for the convenience of understanding the principles of the present application, and are not intended to limit the technical solutions provided by the embodiments of the present application.

Next, a vehicle collision warning method provided in an embodiment of the present application will be described with reference to the drawings.

Referring to fig. 2, the figure is a schematic flow chart of a vehicle collision warning method provided in an embodiment of the present application, and the method may be applied to a roadside unit 101 or an on-board unit 102, and specifically, the method may include the following steps:

s201: the method comprises the steps of obtaining a real-time driving state of a target vehicle and a mapping relation between a historical driving state and a prior collision threat coefficient, wherein the prior collision threat coefficient reflects collision possibility of a plurality of training vehicles in the historical driving state.

In the present embodiment, S201 includes two sub-steps, one of which is to obtain a real-time driving status of the target vehicle, and the other is to obtain a mapping relationship between a historical driving status and an a priori collision threat coefficient. The mapping relation between the historical driving state and the prior collision threat coefficient can be obtained according to the historical driving states of a plurality of training vehicles and the historical collision results, and the mapping relation reflects the prior collision possibility. The real-time running state of the target vehicle is the running state acquired under the current running environment, and is used for calculating a posterior collision threat coefficient of the target vehicle when the target vehicle collides, namely the possibility of future collision. By combining the prior collision threat coefficients and the posterior collision threat coefficients, the prediction of the collision probability of the target vehicle can be made more accurate.

In this embodiment, the real-time driving state of the target vehicle may include a relative speed, a relative distance, and a relative driving direction between the target vehicle and other vehicles around the target vehicle in the current driving environment.

Specifically, the relative speed between the target vehicle and other vehicles around the target vehicle may be a speed of the target vehicle with the vehicles around the target vehicle as a reference, the relative distance between the target vehicle and other vehicles around the target vehicle may be a link distance between the target vehicle and the vehicles around the target vehicle, and the relative driving direction between the target vehicle and other vehicles around the target vehicle may be represented as an included angle between the heading angle of the target vehicle and the heading angles of other vehicles around the target vehicle.

It is understood that the historical driving state may include relative speed, relative distance, and relative driving direction between a number of training vehicles under the historical driving environment. The training vehicles may be training vehicles adjacent to each other at the same time and space, for example, the training vehicles are all at the same intersection at the same time.

The prior collision threat coefficient can be a discrete index and is used for evaluating the possibility of collision of a plurality of training vehicles in a historical driving state. The greater the prior collision threat coefficient is, the greater the possibility that the training vehicles collide in the historical driving state is, and on the contrary, the less the possibility that the training vehicles collide in the historical driving state is.

The prior collision threat coefficient can reflect the collision possibility of a plurality of training vehicles in the historical driving state. Therefore, in order to obtain the mapping relationship between the historical driving state and the prior collision threat coefficients, the prior collision threat coefficients of a plurality of training vehicles in the historical driving state need to be obtained first. The specific method for obtaining the mapping relationship between the historical driving state and the prior collision threat coefficient will be described in detail later.

S202: and obtaining the prior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle and the mapping relation.

In order to obtain the prior collision threat coefficient of the target vehicle, a historical driving state that is the same as or similar to the real-time driving state may be determined according to the real-time driving state of the target vehicle, and it should be noted that the historical driving environment corresponding to the historical driving state may be the same as the current driving environment corresponding to the real-time driving state, for example, the historical driving state and the real-time driving state may be collected at the same intersection, or the historical driving environment corresponding to the historical driving state may be similar to the current driving environment corresponding to the real-time driving state, for example, the historical driving state and the real-time driving state may be collected at intersections at two different locations, respectively.

Then, a prior collision threat coefficient can be obtained according to the historical driving state and the mapping relation between the historical driving state and the prior collision threat coefficient, and the prior collision threat coefficient can be used as the prior collision threat coefficient of the target vehicle. It should be noted that, a specific implementation manner of obtaining the prior collision threat coefficient will be described in detail later.

S203: and calculating the posterior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle.

The posterior collision threat coefficient can reflect the possibility that the target vehicle collides with other vehicles around the target vehicle in the real-time driving state. The greater the posterior collision threat coefficient is, the greater the possibility that the target vehicle collides with the other vehicles around the current vehicle in the real-time traveling state is, and on the contrary, the lesser the possibility that the target vehicle collides with the other vehicles around the current vehicle in the real-time traveling state is.

First, after acquiring the real-time driving state of the target vehicle, that is, acquiring the relative speed, the relative distance, and the relative driving direction between the target vehicle and other vehicles around the target vehicle in the current driving environment, the collision time between the target vehicle and other vehicles around the target vehicle can be obtained according to the relative speed, the relative distance, and the relative driving direction between the target vehicle and other vehicles around the target vehicle.

The collision time of the target vehicle and other vehicles around the target vehicle may be a time period from a current time to a future preset time, where the future preset time is a time when the target vehicle collides with other vehicles around the target vehicle. The shorter the collision time, the greater the possibility that the target vehicle collides with other vehicles around the target vehicle, and conversely, the smaller the possibility that the target vehicle collides with other vehicles around the target vehicle.

The manner in which the time to collision is predicted will be described in detail later.

Secondly, after the collision time of the target vehicle and other vehicles around the target vehicle is predicted, the posterior collision threat coefficient of the target vehicle can be obtained according to the collision time.

Specifically, if the collision time is greater than or equal to a preset threshold, the possibility of the target vehicle colliding with other vehicles around may be considered to be low, and at this time, the posterior collision threat coefficient D of the target vehicle may be 0.

If the collision time is less than the preset threshold, it may be considered that the probability of collision between the target vehicle and other surrounding vehicles is high, and at this time, the posterior collision threat coefficient of the target vehicle may be calculated by using the collision time. For example, the posterior collision threat coefficient for the target vehicle may be calculated using the following formula:

d ═ 1/lgt, where D represents the a posteriori collision threat coefficient and t represents the time of collision of the target vehicle with other vehicles in its current surroundings.

For example, assume that the relative speed between the target vehicle and a vehicle in its current surroundings is 20m/s, the relative distance is 100m, and the relative direction of travel is 0 degrees, and the preset threshold value is 10 s. The collision time t between the target vehicle and the vehicle is calculated to be 5s according to the real-time running state of the target vehicle, and the posterior collision threat coefficient D of the target vehicle colliding with the surrounding vehicle is 1.42(1/lg5) because the collision time t is less than the preset threshold value 10 s.

It should be noted that the preset threshold may be obtained according to the historical driving state, and a specific implementation manner of obtaining the preset threshold will be described in detail later. In one possible embodiment, when the collision time of the target vehicle and other vehicles around the target vehicle is predicted to be greater than the preset threshold, since the possibility of collision between the target vehicle and other vehicles around the target vehicle can be considered to be extremely low, it may not be necessary to calculate the posterior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle, thereby reducing the calculation amount in the whole early warning process.

S204: and early warning is carried out on the target vehicle according to the prior collision threat coefficient and the posterior collision threat coefficient.

Since the real-time running state of the target vehicle at a certain moment may not represent the running tendency of the target vehicle, the collision time obtained according to the real-time running state of the target vehicle may be too small, and an error may exist if the possibility of collision of the vehicle is determined only by the collision time.

In the embodiment, not only the posterior collision threat coefficient which is obtained according to the real-time driving state and reflects the collision possibility of the target vehicle is referred, but also the prior collision coefficient which corresponds to the historical driving state which is the same as or similar to the real-time driving state is referred, and the prior collision coefficient is obtained according to the historical collision result, so that the combination of the prior collision threat coefficient and the posterior collision threat coefficient can more accurately predict the collision possibility of the target vehicle and other surrounding vehicles, and the early warning error rate of the target vehicle is reduced.

In an implementation manner of the embodiment of the present application, a comprehensive threat collision probability may be obtained according to the prior collision threat coefficient and the posterior collision threat coefficient, for example, the comprehensive threat collision probability may be obtained by using the prior collision threat coefficient and the posterior collision threat coefficient and calculating based on bayesian theorem, and specifically, the comprehensive threat collision probability of the target vehicle may be calculated by using the following formula based on bayesian theorem:

wherein, X represents the comprehensive threat collision probability, D represents the posterior collision threat coefficient, P represents the prior collision threat coefficient, and X represents the dot product.

The larger the comprehensive threat collision probability X obtained according to the formula is, the larger the probability that the target vehicle collides is, and on the contrary, the smaller the probability that the target vehicle collides is. If the comprehensive threat collision probability is smaller than a first threshold value, the target vehicle can be considered to have smaller collision threat, and early warning is not needed to be carried out on the target vehicle; if the vehicle speed is greater than the first threshold and less than or equal to the second threshold, the target vehicle is considered to have a light collision threat, and light early warning is carried out on the target vehicle; if the vehicle collision risk is larger than the second threshold and smaller than or equal to the third threshold, the target vehicle is considered to have a moderate collision threat, and moderate early warning is carried out on the target vehicle; if the current time is greater than the third threshold value, the target vehicle is considered to have a severe collision threat, and a severe early warning is given to the target vehicle. The first threshold, the second threshold and the third threshold may be preset by the user according to historical experience, for example, the first threshold may be 0.3, the second threshold may be 0.5, and the third threshold may be 0.7.

In this embodiment of the application, in order to obtain the mapping relationship between the historical driving state and the prior collision threat coefficient, the step of "obtaining the mapping relationship between the historical driving state and the prior collision threat coefficient" in S201 may specifically include the following steps:

step A: and acquiring historical driving states and historical collision results of the training vehicles.

In this embodiment, the historical driving states and the historical collision results of the training vehicles may be obtained by collecting passing vehicles by using a DSRC communication technology, may be directly obtained from a related organization, or may be collected in a simulated vehicle collision experiment. The manner in which the historical driving state and the historical collision result are obtained is not limited.

It should be noted that the plurality of training vehicles may include two training vehicles that are adjacent to each other at the same time and space. For convenience of description, the two training vehicles may be referred to as a first training vehicle and a second training vehicle, respectively, for example, two training vehicles at the same intersection may be referred to as a first training vehicle and a second training vehicle, respectively, at a certain time.

The historical travel state may include a relative speed, a relative distance, and a relative travel direction between the first training vehicle and the second training vehicle under the historical travel environment.

It will be appreciated that the relative speed between the first training vehicle and the second training vehicle may be the speed of the first training vehicle with the second training vehicle as a reference, the relative distance between the first training vehicle and the second training vehicle may be the link distance between the first training vehicle and the second training vehicle, and the relative direction of travel between the first training vehicle and the second training vehicle may be represented by the angle between the heading angle of the first training vehicle and the heading angle of the second training vehicle.

And the historical collision result of each training vehicle may be a collision result of the training vehicle in a historical driving state, and the historical collision result may include one of: crash and no airbag ejection, crash and airbag ejection, no crash and emergency braking, no crash and deceleration to 0 by braking and no crash and no deceleration to 0 by braking.

And B: and obtaining a prior collision threat coefficient in the historical driving state according to the historical driving states and the historical collision results of the training vehicles, so as to obtain a mapping relation between the historical driving states and the prior collision threat coefficient.

Since the historical driving state of the training vehicle changes during the driving process of the training vehicle, in order to obtain the prior collision threat coefficients corresponding to the more accurate historical driving state, the prior collision threat coefficients of the historical driving state under different conditions need to be obtained, and therefore the historical driving state can be divided into multiple conditions.

Taking the first training vehicle and the second training vehicle mentioned in step a as an example, the relative speed between the first training vehicle and the second training vehicle may be divided into a plurality of cases, specifically, a section a (0,20 m/s), a section B (20m/s, 25 m/s), a section C (25m/s, 28 m/s), a section D (28m/s, 30 m/s), a section E (30m/s, ∞), the relative distance between the first training vehicle and the second training vehicle may be divided into a plurality of cases, specifically, a section a (0, 40 m), a section B (40m, 60 m), a section C (60m, 80 m), a section D (80m, 100 m), a section E (100m, ∞), the relative travel direction between the first training vehicle and the second training vehicle may be divided into a plurality of cases, the angle in the traveling direction may be specifically a section 1(0, 30 °), a section 2(30 °, 60 °), a section 3(60 °, 90 °), and a section 4(90 °, 180 °).

Because each training vehicle corresponds to the same historical collision result under different conditions of the historical driving state, the historical collision results corresponding to a plurality of training vehicles under different conditions of the historical driving state can be obtained firstly.

It should be noted that, by analyzing the historical collision result, it can be known that, if the collision result of a training vehicle is a collision and the airbag is not ejected, the prior collision threat coefficient corresponding to the collision result is 1; if the collision result of a training vehicle is collision and an air bag is ejected, the prior collision threat coefficient corresponding to the collision result is 1; if the collision result of a training vehicle is no collision and emergency braking is carried out, the prior collision threat coefficient corresponding to the collision result is 0.9; if the collision result of a training vehicle is not collision and is decelerated to 0 through braking, the prior collision threat coefficient corresponding to the collision result is 0.5; if the collision result of a training vehicle is no collision and the brake is not decelerated to 0, the prior collision threat coefficient corresponding to the collision result is 0.

Therefore, the prior collision threat coefficients corresponding to the historical collision results of the training vehicle under different conditions of the historical driving state can be used as training set samples under different conditions of the historical driving state. That is to say, the training set samples of the historical driving states under different conditions each include a priori collision threat coefficients corresponding to a plurality of historical collision results.

Continuing with the example of the first training vehicle and the second training vehicle described above, it is assumed that the historical driving state includes that under the historical driving environment, the relative speed between the first training vehicle and the second training vehicle is 23m/s, the relative distance is 50m, and the relative driving direction is 20 °, and the collision result is no collision and is decelerated to 0 by braking, i.e., the collision result corresponds to an a priori collision threat coefficient of 0.5, and when the speeds of the first and second training vehicles are both 0, the relative distance is 10 m.

Since the relative speed gradually decreases to 0 during deceleration of the training vehicle, the relative distance gradually decreases to 10m, and the relative driving direction does not change. Therefore, 0.5 of the prior collision threat coefficient may be used as training sample data when the relative speeds of the first training vehicle and the second training vehicle are respectively the section a (0,20 m/s) and the section B (20m/s and 25 m/s), as training sample data when the relative distances are respectively the section a (0, 40 m) and the section B (40m and 60 m), and as training sample data when the relative traveling direction is the section 1(0, 30 °).

After training sample data of the historical driving state under different conditions are acquired, the prior collision threat coefficients of the historical driving state under different conditions can be obtained according to the prior collision threat coefficients corresponding to a plurality of historical collision results in the training sample data, for example, the prior collision threat coefficients corresponding to the historical driving state under different conditions can be calculated by using the following formula based on bayesian theorem:

wherein, Y_jRepresenting the jth historical driving state, x_iThe ith training sample data representing the jth historical driving state, I representing the number of training sample data of the jth historical driving state, P (X | Y)_j) A corresponding a priori crash threat coefficients representing a jth historical driving state.

After the prior collision threat coefficients corresponding to the historical driving states under different conditions are obtained, the prior collision threat coefficients corresponding to the historical driving states can be obtained according to the prior collision threat coefficients corresponding to the historical driving states under different conditions, and the mapping relation between the historical driving states and the prior collision threat coefficients is obtained. For example, since the historical driving states are independent of each other, the prior collision threat coefficient corresponding to the historical driving state can be calculated by using the following formula based on bayesian theorem:

P(X|Y₁,Y₂,…,Y_j)＝P(X|Y₁)P(X|Y₂)…P(X|Y_j) Wherein, P (X | Y)₁,Y₂,…,Y_i) A priori collision threat coefficient, Y, corresponding to the historical driving state_jThe jth case representing the historical driving state, P (X | Y)_j) And the prior collision threat coefficient represents the j-th condition, and j represents the number of different conditions of the historical driving state.

After the mapping relation between the historical driving state and the prior collision threat coefficient is obtained, the prior collision threat coefficient can be obtained according to the historical driving state and the mapping relation between the historical driving state and the prior collision threat coefficient. Next, how to obtain the a priori collision threat coefficient according to the historical driving state will be specifically described.

The historical driving state of each training vehicle corresponds to the same historical collision result and the prior collision threat coefficient corresponding to the historical collision result. Therefore, after the historical driving state, that is, the relative speed, the relative distance and the relative driving direction, of a training vehicle are obtained, the corresponding sections can be determined according to the relative speed, the relative distance and the relative driving direction, and the historical collision result of the training vehicle in the historical driving state and the prior collision threat coefficient corresponding to the historical collision result can be determined according to the sections.

Continuing with the example of the first training vehicle and the second training vehicle, a mapping relationship between the historical driving state corresponding to the first training vehicle and the prior collision threat coefficient is established. Therefore, when the historical traveling states of the first training vehicle and the other training vehicles are obtained as the relative speed of 15m/s, the relative distance of 20m, and the relative traveling direction of 15 °, the section corresponding to the relative speed of 15m/s is determined as the section A (0,20 m/s), the section corresponding to the relative distance of 20m is determined as the section a (0, 40 m), the section corresponding to the relative traveling direction of 15 ° is determined as the section 1(0, 30 ° ], and then, the corresponding one of the historical collision results and the prior collision threat coefficient corresponding to the historical collision result, which are the relative speeds of the first training vehicle and the other training vehicles at 15m/s, the relative distance of 20m, and the relative traveling direction of 15 ° are determined from the sections A (0,20 m/s), the section a (0, 40 m), and the section 1(0, 30 ° ] Historical collision results at a relative distance of 20m and a relative direction of travel of 30 °.

Next, how to obtain the preset threshold value according to the history of the running state will be described.

The historical driving state of the training vehicle can be divided into a plurality of cases, each case corresponds to one collision time, and each case corresponds to one prior collision threat coefficient. Thus, each collision time also corresponds to an a priori threat collision coefficient.

In this embodiment, the collision time of the vehicle may be divided into a plurality of cases, for example, the collision time may be divided into a first interval (0,2 s), a second interval (2s, 5 s), a third interval (5m/s, 8 s), a fourth interval (8s, 10 s), and a fifth interval (10s, infinity).

When the prior collision threat coefficient is 0, it can be considered that the possibility of the vehicle collision is extremely low, and therefore, a collision time can be determined in all collision time intervals with the prior collision threat coefficient being 0, and the collision time is taken as a preset threshold value, for example, the collision time with the shortest time in the collision time intervals.

The possibility of collision of the vehicle in the history traveling state is also affected due to the difference in the history traveling environment. For example, when the historical driving environment includes a weather condition, if the weather condition is clear, the visual sight of the driver is good, and the condition in front of the vehicle can be quickly judged, so that the driver can judge whether braking is needed in advance; if the weather conditions are rain, snow and haze, the visual sight range of the driver is small, the front condition of the vehicle is difficult to judge, and the driver cannot judge whether braking is needed in advance, so that the possibility of collision of the vehicle is increased.

For another example, when the historical driving environment includes a road shape, if the road shape is a straight line, the driver can easily observe the situation in front of the vehicle, so that the driver can judge whether braking is needed in advance; if the road shape is a curve or a cross, the driver cannot observe the front condition of the vehicle in advance, so that the driver cannot judge whether braking is needed in advance, and the possibility of collision of the vehicle is increased.

For another example, when the historical driving environment includes road surface conditions, since friction coefficients of different road surface conditions (such as mountain roads, painted roads and wet roads) are different, the time period for the vehicle to decelerate to 0 by braking is different under different road surface conditions, thereby affecting the possibility of collision of the vehicle.

Therefore, in order to obtain a more accurate collision threat coefficient that matches the actual situation, in one embodiment of the present application, a collision threat coefficient in a history driving state may be obtained according to the history driving state, history collision result, and history driving environment of a plurality of training vehicles.

Specifically, after the prior collision threat coefficients of the historical driving state under different conditions and the prior collision threat coefficients of various historical driving environments are obtained, the prior collision threat coefficients corresponding to the historical driving state can be obtained according to the prior collision threat coefficients of the historical driving state under different conditions and the prior collision threat coefficients of various historical driving environments, and the mapping relation between the historical driving state and the prior collision threat coefficients is obtained. For example, since the historical driving conditions are independent of each other given the historical driving environment, the prior collision threat coefficient corresponding to the historical driving conditions can be calculated by using the following formula based on bayesian theorem:

P(X|Z₁,…,Z_m,Y₁,Y₂,…,Y_j)＝P(X|Z₁,…,Z_m,Y₁)P(X|Z₁,…,Z_m,Y₂)…P(X|Z₁,…,Z_m,Y_j) Wherein, P (X | Z)₁,…,Z_m,Y₁,Y₂,…,Y_j) A priori collision threat coefficient, Y, corresponding to the historical driving state_jJ-th case, Z, representing a historical driving situation_mRepresents the m-th history of travel environment, P (X | Z)₁,…,Z_m,Y_j) Is represented at a given Z₁,…Z_mA priori collision threat coefficients for a jth instance of the historical driving conditions under the condition, j representing a number of different instances of the historical driving conditions, and m representing a number of various historical driving environment instances.

Next, a manner of obtaining the posterior collision threat coefficient will be specifically described. In one implementation manner of the embodiment of the present application, the posterior collision threat coefficient of the target vehicle may be obtained by predicting the collision time of the target vehicle and other vehicles around the target vehicle, wherein the real-time driving state of the target vehicle may include real-time position information of the target vehicle, and accordingly, the step of "calculating the posterior collision threat coefficient of the target vehicle according to the real-time driving state of the target vehicle" in S203 may specifically include the following steps:

predicting the collision time of the target vehicle and other surrounding vehicles according to the real-time position information of the target vehicle and the real-time position information of other surrounding vehicles of the target vehicle;

step b: and obtaining the posterior collision threat coefficient according to the collision time.

The specific implementation manner of step b is the same as that in S203, please refer to the related description in S203, which is not described herein again.

The implementation of step a is described in detail below.

In this embodiment, the time of collision between the target vehicle and other surrounding vehicles may be understood as a time period from a current time to a future preset time, where the future preset time is a time when the real-time position information of the target vehicle overlaps with the real-time position information of other surrounding vehicles.

The real-time location information may be longitude and latitude information of the vehicle, for example, the real-time location information of the target vehicle may be (lat1, lon1), where lat1 is the longitude of the target vehicle and lon1 is the latitude of the target vehicle.

The longitude and latitude information of the target vehicle and the longitude and latitude information of other surrounding vehicles can reflect the position relationship of the target vehicle and other surrounding vehicles. Therefore, it is possible to determine whether or not the latitude and longitude information of the target vehicle overlaps with the latitude and longitude information of other vehicles around the target vehicle. If so, the target vehicle can be considered to be collided with other surrounding vehicles, and the collision moment is taken as a future preset moment, so that the collision time of the target vehicle and other surrounding vehicles can be predicted; if not, the target vehicle and other surrounding vehicles can be considered not to be collided.

However, the real-time position information of the vehicle is only one coordinate information and cannot accurately reflect the actual size of the vehicle, for example, when two vehicles have collided, but the real-time position information of the two vehicles does not overlap.

In order to accurately predict the future preset time when the target vehicle collides with other vehicles around, a safety zone may be preset for the target vehicle and other vehicles around based on the real-time position information of the target vehicle and the real-time position information of the other vehicles around. The safety area may be an area including real-time position information of the vehicle, for example, a safety area having a length of 5 meters and a width of 2 meters may be set with the coordinate information of the target vehicle as a midpoint. In this way, the collision time of the target vehicle with other surrounding vehicles can be determined by predicting a future preset time at which the safety zone of the target vehicle overlaps with the safety zones of other surrounding vehicles.

Therefore, in an implementation manner of this embodiment, the step a may include the following steps:

step a 1: predicting the position of the target vehicle at a future preset time according to the real-time position information of the target vehicle, and determining a safety area of the target vehicle based on the position of the target vehicle at the future preset time;

step a 2: predicting the arrival positions of other surrounding vehicles at the same future preset time according to the real-time position information of the other surrounding vehicles, and determining safety regions of the other surrounding vehicles based on the arrival positions of the other surrounding vehicles at the future preset time;

step a 3: if the safety zone of the target vehicle overlaps with the safety zones of the other vehicles around, the time period from the current time to the future preset time may be determined as the time of collision of the target vehicle with the other vehicles around.

Since the method of predicting the arrival position of the target vehicle at the preset time in the future is the same as the method of predicting the arrival positions of other vehicles around the preset time in the future, the specific implementation manner of step a1 and step a2 is the same. Therefore, the following will take the target vehicle in step a1 as an example, and a method for predicting the position where the target vehicle arrives at the future preset time will be specifically described, and step a2 may refer to the related description of step a1, and will not be described herein again.

In this embodiment, the real-time running state of the target vehicle may further include: speed information of the target vehicle (such as a current speed, a current acceleration of the target vehicle), yaw information (such as a heading angle, a yaw rate of the target vehicle), and current running environment information (such as weather conditions, road shape, road surface conditions, ground friction coefficient).

Firstly, after the real-time position information of the target vehicle is acquired, since the real-time position information is longitude and latitude information of the target vehicle, for convenience of calculation, the longitude and latitude information can be converted into coordinate information which is convenient for calculation, for example, coordinate values of a plane rectangular coordinate system.

Specifically, a road position around the target vehicle may be used as the origin of the rectangular plane coordinate system, such as a midpoint of an intersection, a middle position between the target vehicle and other vehicles around, and the like. Then, based on the positional relationship between the latitude and longitude information (lat1, lon1) of the target vehicle and the latitude and longitude information (lat2, lon2) of the road position, the coordinate values (x) of the target vehicle in a rectangular plane coordinate system with the road position as the origin can be obtained₁，y₁) And converting the course angle theta of the target vehicle into the coordinate value (x)₁，y₁) The included angle theta between the X-axis positive half shaft and the X-axis positive half shaft in the plane rectangular coordinate system_x。

Since the target vehicle is in different motion states, such as straight going and turning, the variation trend of the real-time position information of the target vehicle is different, i.e. the variation trend of the coordinate values of the target vehicle in the rectangular plane coordinate system is different. Therefore, before predicting the position where the target vehicle arrives at the future preset time, it is necessary to determine the motion state of the target vehicle.

In one implementation manner of the embodiment, the motion state of the target vehicle may be determined according to the curvature radius of the target vehicle, for example, the curvature radius of the target vehicle may be calculated by the following formula:

r＝(V/γ)

where r represents the current radius of curvature of the target vehicle, V represents the current speed of the target vehicle, and γ represents the yaw rate of the target vehicle.

It should be noted that when the moving state of the target vehicle is a straight line, the target vehicle may be considered to be moving around the earth. Since the elevation of each location on the earth's surface is different, the radius of curvature of the target vehicle should be greater than or equal to the radius of curvature (r) of the earth when the target vehicle travels straight_Ground32767), i.e., the radius of curvature of the target vehicle should be greater than or equal to 32767.

When the moving state of the target vehicle is turning, it can be considered that the target vehicle is turning around a certain ground position. Since the center of the circle surrounded by the target vehicle is the ground position, the distance from the center of the circle of the target vehicle can be smaller than the radius of the earth. Therefore, when the target vehicle turns, the curvature radius of the target vehicle should be smaller than the curvature radius of the earth, i.e. the curvature radius of the target vehicle should be smaller than 32767.

Secondly, after the motion state of the target vehicle is determined, the position and the safety area reached by the target vehicle at the future preset time can be predicted according to the real-time running state of the target vehicle.

Specifically, when the target vehicle is traveling in a straight line, the position where the target vehicle arrives at a future preset time can be predicted from the real-time position information of the target vehicle, the current speed, the current acceleration, the yaw angle, and the ground friction coefficient.

The displacement of the target vehicle at the preset time in the future may be determined according to the current speed, the current acceleration and the ground friction coefficient of the target vehicle, and may be calculated by the following formula:

s＝v*t+0.5*(a-fg)*t²

wherein t represents a future preset time; s represents the displacement of the target vehicle at a future preset time t; v represents the current speed of the target vehicle; a represents the current acceleration of the target vehicle; g represents the acceleration of gravity; f represents a current road friction coefficient, which can be obtained from current running environment information such as weather conditions, road shape, road surface conditions.

It should be noted that, because different weather conditions may cause different humidity conditions of the road surface, the weather conditions may be represented by the humidity conditions of the road surface, for example, when the weather conditions are rainy days, the humidity conditions of the road surface may be wet, and when the weather conditions are sunny days, the humidity conditions of the road surface may be dry; the road surface condition may generally include a road surface condition and a road surface service time, specifically, the road surface condition may be a material of the road surface, such as asphalt, gravel or concrete, and generally, the longer the road surface service time is, the lower the road friction coefficient is. The correspondence between the road friction coefficient and the road surface condition, the road surface humidity condition, and the road surface use time can be obtained from table 1.

TABLE 1

Road surface condition	Dry or wet	Service time of road surface	Coefficient of road friction
				Asphalt	Drying	New road	0.85
Asphalt	Drying	One to three years	0.75
				Asphalt	Drying	More than three years	0.70
Asphalt	Moisture content	New road	0.80
				Asphalt	Moisture content	One to three years	0.65
Asphalt	Moisture content	More than three years	0.60
				Concrete and its production method	Drying	New road	0.90
Concrete and its production method	Drying	One to three years	0.78
				Concrete and its production method	Drying	More than three years	0.70
Concrete and its production method	Moisture content	New road	0.78
				Concrete and its production method	Moisture content	One to three years	0.70
Concrete and its production method	Moisture content	More than three years	0.62
				Sandstone	Drying	Ten years or less	0.68

After determining the displacement of the target vehicle at the future preset time, the position of the target vehicle at the future preset time may be determined according to the displacement and the real-time position information of the target vehicle, for example, the position of the target vehicle at the future preset time may be calculated by the following formula:

(x_t,y_t)＝(x₁-s*cosθ_x,y₁-s*sinθ_x)

wherein t represents a future preset time; s represents the displacement of the target vehicle at a future preset time t; x is the number of_tAn abscissa value representing the target vehicle at the future preset time t; y is_tRepresenting the longitudinal coordinate value of the target vehicle at the future preset time t; x is the number of₁An abscissa value representing the target vehicle at the present time; y is₁A vertical coordinate representing the target vehicle at the current momentA value; theta_xRepresents the coordinate value (x)_t,y_t) And the included angle between the X-axis positive half shaft and the X-axis positive half shaft in the plane rectangular coordinate system.

Then, a safety zone of the target vehicle at the future preset time may be obtained according to the position of the target vehicle at the future preset time, for example, the safety zone may be a zone surrounded by the following four coordinate values:

(x_t+A/2,y_t+B/2)，(x_t+A/2,y_t-B/2)，(x_t-A/2,y_t+B/2)，(x_t-A/2,y_t-B/2)

wherein t represents a future preset time; x is the number of_tAn abscissa value representing the target vehicle at the future preset time t; y is_tRepresenting the longitudinal coordinate value of the target vehicle at the future preset time t; a represents the width of a preset safety region; b represents the length of the preset safety zone.

When the target vehicle runs in a turn, the position where the target vehicle arrives at the future preset time can be predicted according to the real-time position information, the current speed, the current acceleration, the yaw angle, the yaw angular velocity and the ground friction coefficient of the target vehicle.

The displacement of the target vehicle at the future preset time and the projection of the displacement in the direction of X, Y axis may be determined according to the current speed, the current acceleration and the ground friction coefficient of the target vehicle, for example, the displacement of the target vehicle at the future preset time and the projection thereof in the direction of X, Y axis may be calculated by the following 4 formulas:

s＝v*t+0.5*(a-fg)*t²，arc_angle＝s/r，

arc_liner_x＝r*sin(arc_angle)，arc_liner_y＝r*cos(arc_angle)

wherein t represents a future preset time; s represents the displacement of the target vehicle at a future preset time t; v represents the current speed of the target vehicle; a represents the current acceleration of the target vehicle; g represents the acceleration of gravity; f represents a current ground friction coefficient, which can be obtained from table 1 depending on current running environment information such as weather conditions, road shape, road surface conditions; r represents the current radius of curvature of the target vehicle; arc _ angle represents the rotation angle of the target vehicle at a preset time t in the future; arc _ line _ X represents the projection of the displacement s in the X-axis direction; arc _ line _ Y represents the projection of the displacement s in the Y-axis direction.

After determining the displacement of the target vehicle at the future preset time and the projection of the displacement in the X, Y axis direction, the position of the target vehicle at the future preset time may be determined according to the displacement, the real-time position information of the target vehicle and the projection of the displacement in the X, Y axis direction, for example, the position of the target vehicle at the future preset time may be calculated by the following formula:

x_t＝x₁+(arc_liner_x*cos(θ_x)-y₁*sin(θ_x))，

y_t＝y₁+(arc_liner_x*sin(θ_x)+arc_liner_y*cos(θ_x))

wherein t represents a future preset time; x is the number of_tAn abscissa value representing the target vehicle at the future preset time t; y is_tRepresenting the longitudinal coordinate value of the target vehicle at the future preset time t; x is the number of₁An abscissa value representing the target vehicle at the present time; y is₁A longitudinal coordinate value representing the target vehicle at the current moment; theta_xRepresents the coordinate value (x)_t,y_t) The included angle between the X-axis positive half shaft and the X-axis positive half shaft in the plane rectangular coordinate system; arc _ line _ X represents the projection of the displacement s in the X-axis direction; arc _ line _ Y represents the projection of the displacement s in the Y-axis direction.

(x_t+A/2,y_t+B/2)，(x_t+A/2,y_t-B/2)，(x_t-A/2,y_t+B/2)，(x_t-A/2,y_t-B/2)

wherein t represents a future preset time; x is the number of_tRepresenting the objectThe abscissa value of the vehicle at the future preset time t; y is_tRepresenting the longitudinal coordinate value of the target vehicle at the future preset time t; a represents the width of a preset safety region; b represents the length of the preset safety zone.

The implementation of step a3 is described in detail below.

By executing steps a1 and a2, the position and the safe zone where the target vehicle arrives at the future preset time, and the positions and the safe zones where other vehicles around the target vehicle arrive can be obtained.

Because the safety regions of the target vehicle and other surrounding vehicles are surrounded by four line segments, if any one line segment in the safety region of the target vehicle is intersected with any one line segment in the safety regions of the other surrounding vehicles, the future preset time corresponding to the safety region can be considered as the collision time of the target vehicle and the other surrounding vehicles. At this time, the time period from the present time to the future preset time may be determined as the time of collision of the target vehicle with the other surrounding vehicles.

How to judge the intersection of two line segments will be illustrated in conjunction with fig. 3.

As shown in fig. 3, taking a segment A, B as an example, two ends of the segment a are respectively a point P1 and a point P2, and two ends of the segment B are respectively a point P3 and a point P4.

First, a point P3 may be connected to a point P1 and a point P2 to form a line segment C, D, wherein a vector v1 of the line segment C is from P3 to P1, a vector v2 of the line segment D is from P3 to P2, and a vector v3 of the line segment B is from P3 to P4.

Then, whether the line segment A, B intersects can be judged according to the vectors v1, v2 and v3, for example, whether A, B intersects can be judged by the following formula:

Z＝(v1×v3)*(v2×v3)

if Z is less than or equal to 0, then line segment A, B is intersected, otherwise line segment A, B is not intersected.

Based on the vehicle collision early warning method provided by the embodiment, the embodiment of the application also provides a vehicle collision early warning device, and the working principle of the vehicle collision early warning device is explained in detail by combining the attached drawings.

Referring to fig. 4, the figure is a block diagram of a vehicle collision warning apparatus according to an embodiment of the present application.

The vehicle collision warning apparatus provided by the present embodiment may include:

a first obtaining unit 401, configured to obtain a real-time driving state of a target vehicle and a mapping relationship between a historical driving state and a priori collision threat coefficient, where the priori collision threat coefficient reflects collision possibility of a plurality of training vehicles in the historical driving state;

a second obtaining unit 402, configured to obtain a priori collision threat coefficient of the target vehicle according to a real-time driving state of the target vehicle and the mapping relationship;

a calculating unit 403, configured to calculate a posterior collision threat coefficient of the target vehicle according to a real-time driving state of the target vehicle, where the posterior collision threat coefficient reflects a possibility that the target vehicle collides with other vehicles around the target vehicle;

and the early warning unit 404 is used for early warning the target vehicle according to the prior collision threat coefficient and the posterior collision threat coefficient.

Optionally, the first obtaining unit 401 includes:

weather conditions, road shape, and road surface conditions.

Optionally, the road shape comprises: straight, turning, or crossing.

Optionally, the historical collision result includes one of the following:

the calculation unit 403 includes:

Optionally, the predictor unit includes:

the first prediction module comprises:

Based on the vehicle collision early warning method and device provided by the above embodiments, the embodiment of the application further provides a vehicle collision early warning device, and the device comprises:

a processor and a memory storing a program;

wherein the processor, when executing the program, performs the following:

When introducing elements of various embodiments of the present application, the articles "a," "an," "the," and "said" are intended to mean that there are one or more of the elements. The terms "comprising," "including," and "having" are intended to be inclusive and mean that there may be additional elements other than the listed elements.

It should be noted that, as one of ordinary skill in the art would understand, all or part of the processes of the above method embodiments may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when executed, the computer program may include the processes of the above method embodiments. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus embodiment, since it is substantially similar to the method embodiment, it is relatively simple to describe, and reference may be made to some descriptions of the method embodiment for relevant points. The above-described apparatus embodiments are merely illustrative, and the units and modules described as separate components may or may not be physically separate. In addition, some or all of the units and modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

The foregoing is directed to embodiments of the present application and it is noted that numerous modifications and adaptations may be made by those skilled in the art without departing from the principles of the present application and are intended to be within the scope of the present application.

Claims

1. A vehicle collision warning method, characterized in that the method comprises:

obtaining a prior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle and the mapping relation; the prior collision threat coefficient of the target vehicle is the collision possibility corresponding to the training vehicle in the historical driving state which is the same as or similar to the real-time driving state;

2. The method of claim 1, wherein obtaining a mapping of historical driving states to a priori collision threat coefficients comprises:

3. The method of claim 2, wherein the number of training vehicles includes a first training vehicle and a second training vehicle;

4. The method according to claim 3, wherein the real-time traveling state of the target vehicle includes a relative speed, a relative distance, and a relative traveling direction between the target vehicle and other surrounding vehicles in the current traveling environment.

5. The method according to claim 4, wherein the historical travel environment and the current travel environment each include at least one of:

weather conditions, road shape, and road surface conditions.

6. The method of claim 5, wherein the road shape comprises: straight, turning, or crossing.

7. The method of claim 2, wherein the historical collision results comprise one of:

8. The method of claim 1, wherein the real-time travel status of the target vehicle includes real-time location information;

9. The method of claim 8, wherein predicting the time of collision between the target vehicle and other vehicles in the vicinity of the target vehicle based on the real-time location information of the target vehicle and the real-time location information of the other vehicles in the vicinity of the target vehicle comprises:

10. The method of claim 9, wherein the real-time driving status of the target vehicle further comprises: the speed information, the yaw information and the current driving environment information of the target vehicle;

11. The method of claim 10, wherein the speed information comprises: the current speed and the current acceleration, the yaw information comprises a yaw angle, and the current driving environment information comprises a ground friction coefficient;

12. The method of claim 10, wherein the speed information comprises: the current speed and the current acceleration, the yaw information comprises a yaw angle and a yaw angular speed, and the current running environment information comprises a ground friction coefficient;

13. A vehicle collision warning apparatus, characterized in that the apparatus comprises:

the second obtaining unit is used for obtaining a prior collision threat coefficient of the target vehicle according to the real-time running state of the target vehicle and the mapping relation; the prior collision threat coefficient of the target vehicle is the collision possibility corresponding to the training vehicle in the historical driving state which is the same as or similar to the real-time driving state;

14. A vehicle collision warning apparatus, characterized in that the apparatus comprises:

a processor and a memory storing a program;

wherein the processor, when executing the program, performs the following: