CN116341288A - A Modeling Method for Safety Field of Heterogeneous Traffic Popular Vehicles - Google Patents

Abstract

The invention belongs to the field of traffic control systems, and in particular relates to a method for modeling the safety field of heterogeneous traffic popular vehicles, respectively constructing a CAV vehicle field model, an HDV vehicle field model, an environmental field model with the The environmental field model steps, and draw a schematic diagram of the driving safety field strength, the present invention is based on the idea of heterogeneous traffic flow differentiation modeling, not only considering the differences in the perception and force of CAV vehicles and HDV vehicles, but also constructing the vehicle field model and The scope constraints of the CAV vehicle field are improved; and considering the influence of HDV driver personality differences and driving environment on driving safety, a comprehensive index of driver environmental psychological visibility is constructed, and the concept of psychological field is used to deduce the HDV driver's performance in mixed driving. The psychological force established the vehicle field model of HDV traditional vehicles.

Description

A Modeling Method for Safety Field of Heterogeneous Traffic Popular Vehicles

technical field

The invention belongs to the field of traffic control systems, and in particular relates to a modeling method for the safety field of popular vehicles in heterogeneous traffic.

Background technique

The inherent chaos of freeway ramp merges is considered to be one of the main causes of traffic accidents and congestion. In the future, connected and autonomous vehicles (CAV) will coexist with human-driven vehicles (HDV) for a long time, which is called heterogeneous traffic flow characteristics, and the car-following and The characteristics and mechanism of micro-traffic behavior such as lane changing are crucial to improving traffic safety and traffic efficiency in mixed traffic systems.

At present, the heterogeneous traffic flow modeling composed of CAV and HDV is mostly based on the use of cellular automata, such as a continuous cellular automata-based vehicle network traffic flow modeling method disclosed in Chinese patent CN115601958 A; Chinese patent CN113204863 B disclosed A manual-CACC self-driving vehicle mixed flow simulation method based on cellular automata; Chinese patent CN 106652564 A discloses a traffic flow cellular automata modeling method under the Internet of Vehicles environment, etc. But the explainability of cellular automata is weak for embodying the safety factors of vehicle operation. In order to highlight the importance of safety for autonomous driving traffic control, a vehicle field model and an environment field model for heterogeneous traffic flow driving behavior based on the safety potential field theory have attracted great attention from the academic community.

Among them, the vehicle field modeling problem based on the safety potential field has been reported in the article "Research on the Movement Situation Model of Mixed Traffic Flow Based on the Safety Potential Field" and so on. The later CAV vehicle field model replaces the HDV vehicle field, ignoring the acceleration, position and other information that cannot be observed by HDV vehicles, and does not consider the driver's personality, vision and trust level of HDV vehicles in heterogeneous traffic flows. Influencing factors, it is impossible to show the degree of influence of different types of drivers on the range of the safety field, and there is no research on the modeling and comprehensive analysis of the safety field of two types of vehicles in heterogeneous traffic flow. At the same time, because the potential field strength in the vehicle field model is limited by the range of action, only a short-range force is applied to the surrounding vehicles within a short distance, and the force changes with the change of vehicle speed, acceleration, distance and other factors. When the target vehicle When the distance from the surrounding vehicles exceeds a certain range, the impact on the field strength of the surrounding vehicles is negligible, but most scholars have not restricted the scope of the field force at present.

As for the environmental field modeling problem based on the safety potential field, the existing reports only use the environmental field modeling with the lateral distance as a variable, without considering the key influence of the longitudinal distance, which cannot reflect the real lane change in some special scenarios The internal mechanism of the vehicle, such as the acceleration lanes in the merge area, the number of lanes reduced due to traffic accidents or temporary construction, etc. The specific performance is: the variable of the existing environmental field is the lateral distance from the vehicle to the road boundary and the marking line, which can better describe the restrictions on the behavior of vehicles by environmental facilities and laws and regulations when there is no change in road conditions. However, when the number of lanes decreases, the vehicle must change lanes to the adjacent lane when the lane is approaching the end. Simulation experiments show that the use of the existing environmental field model will increase the number of vehicles that have not successfully changed lanes, and the merging vehicles that have not successfully changed lanes can only stop at the end of the lane, avoiding peak traffic flow and waiting for the opportunity to merge and change lanes. At this time, the vehicles will suffer Greater driving risk does not match the actual lane change situation.

Contents of the invention

Aiming at the indiscriminate modeling problem existing in the safety field modeling of heterogeneous traffic popular vehicles, the present invention reconstructs the existing vehicle field models of CAV and HDV based on the safety potential field theory, and solves the problem that the existing vehicle field models do not fully describe different types of The degree of influence of the driver on the safe field. Aiming at the problem that the lane change environment does not consider the influence of the longitudinal distance of the lane in some special cases of the existing heterogeneous traffic flow, the present invention establishes a driving environment field with the longitudinal distance as a variable, which truly reflects some special cases, such as merging The impact of reduction in the number of temporary lanes caused by acceleration lanes, road construction or traffic accidents in the district on vehicle operation is consistent with the actual lane change situation. Technical scheme of the present invention is as follows:

A method for modeling the safety field of popular vehicles in heterogeneous traffic, comprising the following steps:

Step A: Construct the CAV vehicle field model

A1 quantify the target vehicle attributes;

A2 determines the pseudo-range;

The degree of danger of the target vehicle also depends on the relative position of the target vehicle and the surrounding potential risk vehicles. Referring to the Euclidean distance, the actual space distance of the potential risk vehicles approaching the target vehicle at different angles is corrected to obtain the pseudo-range;

A3 coordinate transformation

轴方向行驶时，车身会产生一定偏转，车辆场模型也应随之偏转，尤其在换道场景中，公式为：Consider the potential risk that the vehicle will not

When driving in the axial direction, the body will deflect to a certain extent, and the vehicle field model should also deflect accordingly, especially in lane changing scenarios, the formula is:

；

;

为场模型逆时针偏转航向角；It is stipulated that the counterclockwise direction is positive, where:

deflect the heading angle counterclockwise for the field model;

A4 Determining the CAV vehicle field model

；CAV vehicle field strength described in terms of functions in the form of the Yukawa potential used to describe short-range interactions between nucleons

;

A5 Calibrate the parameters of the CAV vehicle field model and constrain its scope of action

Using the differential evolution algorithm, the objective function is to minimize the difference between the force range of the CAV vehicle field and the time headway in the strong car-following state, and the performance parameters of the undetermined coefficients in the established CAV vehicle field model are calibrated to make it The field strength distribution is more in line with the real driving state of the vehicle. Taking the headway of strong car-following as the critical action range of the field strength of the CAV vehicle, the field strength value of the CAV vehicle can be obtained, thereby constraining the action range of the CAV vehicle field;

Step B: Construct HDV vehicle field model

Step B1: Constrain the scope of the HDV vehicle field

The HDV vehicle can maintain a stable state when following the vehicle ahead with a headway of 2.7s. At this time, the field strength of the vehicle is a critical value of 0, namely:

；

;

为标准车头时距；/>

为自车车速；In the formula:

is the standard headway; />

is the vehicle speed;

Step B2: Build the driver's environmental psychological tolerance

Using fuzzy theory, select the typical characteristic factors that reflect the characteristics of the driver: individual experience, environmental visibility, and psychological trust, determine the input eigenvalues of the driver's environmental psychological tolerance evaluation model, and fuzzify the eigenvalues. Through fuzzy rules Logical operation to obtain a fuzzy quantity of driver's environmental psychological tolerance, and then use defuzzification to convert the fuzzy quantity into a precise specific value, which is the driver's environmental psychological tolerance finally calculated by the driver's environmental psychological tolerance evaluation model. Tolerance;

Step B3: Determine the HDV vehicle field model

；

;

为HDV车辆场场强；/>

为与函数极值有关的调整系数； />

为驾驶员环境心理承受度；/>

为安全距离的临界阈值；/>

为与速度相关的待定系数；/>

为伪距离；In the formula:

is the HDV vehicle field strength; />

is the adjustment coefficient related to the extremum of the function; />

For the driver's environmental psychological tolerance; />

is the critical threshold of the safety distance; />

is the undetermined coefficient related to speed; />

is the pseudo-distance;

Step C: Construct an environmental field model with lateral distance as a variable

Select an exponential function to construct an environmental field with the lateral distance as a variable, and make its value infinite at the boundary to prevent the vehicle from leaving;

Step D: Construct an environmental field model with longitudinal distance as a variable

The repulsive potential of the potential field theory is selected to construct the vehicle boundary environment field to quantitatively characterize the impact of the reduction of the number of lanes on vehicle driving safety in the change of road conditions, and to establish an environment field for the lane along the direction of traffic flow;

Step E: Draw a schematic diagram of driving safety field strength

Draw CAV vehicle field, HDV vehicle field, multi-vehicle superposition vehicle field and environment field through MATLAB.

的等效质量/>

的公式为：As a preference of the present invention, in said A1, the target vehicle

The equivalent quality of />

The formula is:

；

;

为目标车辆/>

的等效质量；/>

为目标车辆/>

的实际质量；/>

为目标车辆/>

的速度。In the formula:

for the target vehicle />

The equivalent quality of;/>

for the target vehicle />

the actual quality of

for the target vehicle />

speed.

的公式为：As a preference of the present invention, in said A2, pseudorange

The formula is:

；

;

与/>

分别为车辆长度与宽度，/>

为与道路有关的待定系数。In the formula:

with />

are the length and width of the vehicle, respectively,

is the undetermined coefficient related to the road.

公式为：As a preference of the present invention, in said A4, the field strength of the CAV vehicle

The formula is:

；

;

；

;

、/>

均为待定系数；/>

为目标车辆质心所在空间的位置坐标；/>

为目标车辆当前加速度；/>

为目标车辆周围某点到该车辆质心所在空间坐标/>

的夹角；

为原始坐标偏转后的取值。In the formula:

, />

are undetermined coefficients; />

is the position coordinates of the space where the center of mass of the target vehicle is located; />

is the current acceleration of the target vehicle; />

Space coordinates from a point around the target vehicle to the center of mass of the vehicle />

the included angle;

It is the value after the original coordinate deflection.

As a preference of the present invention, in said A5, the specific method of calibrating the CAV vehicle field model parameters is as follows:

A5.1 Screening the dataset of natural vehicle driving trajectories

A5.2 Preprocessing the dataset

According to the filter conditions of the car-following state, ensure that the front and rear vehicles are driving in the same lane, and set the distance between the front and rear vehicles within the range of 2-150m. When the distance is too small, the default is to eliminate vehicles in queuing state, and when the distance is too large, the default is to eliminate in free flow state; set The minimum car-following time of 10s, if the time length is too long or too small, it will be eliminated by default as the failure to reach a stable state; by sorting out and analyzing the trajectory data of the data set, the distribution interval of each car-following vehicle data is obtained, and the effective follow-up vehicle is extracted by analyzing its distribution law. Chi data;

A5.3 Obtain the time headway under strong car-following state

A5.4 Through the differential evolution algorithm, use the effective car-following data to calibrate the performance parameters of the undetermined coefficients in the CAV vehicle field model; take the headway under the strong car-following state as the critical range of the CAV vehicle field strength, and obtain the CAV The field strength value of the vehicle field constrains the scope of action of the CAV vehicle field.

As a preference of the present invention, in said B2, the specific construction method of the driver's environmental psychological tolerance is as follows:

B2.1 Determine the indicators of environmental psychological tolerance

Determine the driver's individual experience, environmental visibility, and psychological trust as indicators of environmental psychological tolerance;

B2.2 Determine the domain of each indicator

The evaluation model of the driver's environmental psychological tolerance is three-input and single-output. The discourse interval of the individual experience degree is set as [0,5], the fuzzy subset is {Low, Medium, High}, and the discourse interval of the environmental visibility is [0 ,3], the fuzzy subset is {Near, Middle, Far}, the psychological trust domain is [-3,3], the fuzzy subset is {negative large NB, negative small NS, zero ZO, positive small PS, positive large PB};

B2.3 Determine the membership function and membership degree of each index

When the fuzzy input is set as the driver's individual experience degree and environmental visibility, the triangular state membership function is used, and when the fuzzy input is the psychological trust degree, the Gaussian shape membership function is selected, and the membership degree of each index is obtained according to the membership function;

B2.4 Establish the driver's environment psychological tolerance evaluation model rule table according to the degree of membership

B2.5 Determine the relationship between the three input indicators from the rule table, and get the tolerance value in the range of [0,1], which is the psychological tolerance of the driver's environment.

As a preference of the present invention, in step C, the environmental field model formula with the lateral distance as a variable is:

；

;

表示车道外侧道路边界线，过近可与车辆发生实质性碰撞；/>

为车辆行驶坐标；/>

与/>

分别为道路标线与边界风险场的风险系数；/>

为车道总宽度；/>

为势场收敛系数。In the formula: assuming that the vehicle travels on a two-lane road,

Indicates the road boundary line on the outside of the lane, too close to a substantial collision with the vehicle; />

is the coordinates of the vehicle; />

with />

are the risk coefficients of road markings and boundary risk fields respectively; />

is the total width of the lane; />

is the convergence coefficient of the potential field.

As a preference of the present invention, in step D, the environmental field model formula with the longitudinal distance as a variable is:

；

;

为道路标线与边界风险场的风险系数；/>

为车辆/>

到加速车道末端的矢量距离。In the formula:

is the risk coefficient of the road marking and boundary risk field; />

for vehicles />

Vector distance to the end of the acceleration lane.

As a preference of the present invention, in step E, the specific process of the schematic diagram of driving safety field strength is:

，车辆加速度/>

，车辆偏航角/>

，车辆质量/>

，车辆位置坐标，以及用差分进化方法标定出的/>

、/>

、/>

、/>

；输出：CAV车辆场场强/>

；CAV vehicle field, input: vehicle speed

, vehicle acceleration />

, vehicle yaw angle />

, vehicle mass/>

, the vehicle position coordinates, and the calibrated /> by the differential evolution method

, />

, />

, />

;Output: CAV vehicle field strength />

;

、驾驶员环境心理承受度/>

、车辆速度/>

；车辆位置坐标；输出：HDV车辆场场强/>

；HDV vehicle field, input: Adjustment coefficients related to the extrema of the function

, Driver's environmental psychological tolerance />

, vehicle speed />

;Vehicle position coordinates; Output: HDV vehicle field strength/>

;

、/>

，两车加速度/>

、/>

，两车偏航角/>

、/>

，两车质量/>

、/>

，两车位置坐标，以及用差分进化方法标定出的/>

、/>

、/>

、/>

；输出：多车叠加后车辆场场强/>

；Multi-vehicle superimposed vehicle field, input: speed of two vehicles

, />

, the acceleration of the two vehicles />

, />

, the yaw angle of the two vehicles />

, />

, mass of two cars/>

, />

, the position coordinates of the two vehicles, and the /> calibrated by the differential evolution method

, />

, />

, />

;Output: vehicle field strength after multi-vehicle superposition/>

;

，道路标线与边界风险场的风险系数/>

、/>

，势场收敛系数/>

；输出：以横向距离为变量的环境场场强/>

；Environmental field with lateral distance as a variable, input: vehicle coordinates, total lane width

, the risk coefficient of the road marking and boundary risk field />

, />

, potential field convergence coefficient />

;Output: ambient field strength as a variable in lateral distance/>

;

，车辆/>

到加速车道末端的矢量距离/>

；输出：以纵向距离为变量的环境场/>

。Environmental field with longitudinal distance as a variable, input: risk coefficient of road marking and boundary risk field

, vehicle />

vector distance to end of acceleration lane />

;Output: ambient field with longitudinal distance as variable />

.

The beneficial effects of the present invention are as follows:

(1) Based on the idea of differential modeling of heterogeneous traffic flow, the present invention not only considers the differences in perception and force between CAV vehicles and HDV vehicles, but also builds a vehicle field model and improves the constraints on the range of action of CAV vehicle fields; Considering the influence of HDV driver's personality differences and driving environment on driving safety, a comprehensive index of driver's psychological visibility of environment is constructed, and the psychological force of HDV drivers in mixed driving is deduced with the help of the concept of psychological field, and the HDV traditional vehicle is established. vehicle yard model. Combining the determined vehicle field model with the environmental field model representing elements such as road boundaries and markings, a safety field model that jointly describes the driving state and safety of heterogeneous traffic flows is formed;

(2) For the first time, the present invention adds an environmental field model with longitudinal distance as a variable to improve the driving environment field model, which truly reflects the impact of the environmental field on the vehicle's operating conditions, which is consistent with the actual lane change situation, and the situation analysis It shows that the vehicle field model can quantitatively analyze the driving risk of the vehicle, and its contour can visualize the field strength distribution of the vehicle, characterize the vehicle driving safety space, and strengthen the interpretability of the heterogeneous traffic flow safety field model. The dynamic driving risk of vehicles can be quantitatively described, and the car-following model and lane-changing model based on the safety potential field can be derived according to the driving safety field, which lays a theoretical foundation for the study of the motion situation of vehicles under heterogeneous traffic flow.

Description of drawings

Other objects and results of the present invention will become clearer and easier to understand by referring to the following description in conjunction with the accompanying drawings, and with a more comprehensive understanding of the present invention. In the attached picture:

Fig. 1 is a flowchart of the present invention;

；Fig. 2 is a CAV vehicle safety potential field model figure among the present invention,

;

；Fig. 3 is a CAV vehicle safety potential field model figure among the present invention,

;

；Fig. 4 is CAV vehicle safety potential field model figure among the present invention,

;

；Fig. 5 is a CAV vehicle safety potential field model figure among the present invention,

;

Fig. 6 is a model diagram of HDV vehicle safety potential field in the present invention,

Fig. 7 is a schematic diagram of multi-vehicle superposition following and galloping field strength contours in the present invention;

Fig. 8 is a model diagram of the main line lane environment safety potential field in the present invention;

Fig. 9 is a model diagram of the environment safety potential field of the acceleration lane in the present invention.

Detailed ways

In order to enable those skilled in the art to better understand the technical solutions and advantages of the present invention, the application is described in detail below in conjunction with the accompanying drawings, but it is not intended to limit the protection scope of the present invention.

Refer to Figure 1: A modeling method for the safety field of heterogeneous traffic popular vehicles, including the following steps:

A1 quantifies the attributes of the target vehicle, the formula is:

；

;

为目标车辆/>

的等效质量；/>

为目标车辆/>

的实际质量；/>

为目标车辆/>

的速度；In the formula:

for the target vehicle />

The equivalent quality of;/>

for the target vehicle />

the actual quality of

for the target vehicle />

speed;

A2 Determination of pseudorange

的公式为：The degree of danger of the target vehicle also depends on the relative position of the target vehicle and the surrounding potential risk vehicles. Referring to the Euclidean distance, the actual space distance of the potential risk vehicles approaching the target vehicle at different angles is corrected to obtain the pseudo-distance

The formula is:

；

;

为安全距离的临界阈值；/>

为与速度相关的待定系数；/>

与/>

分别为车辆长度与宽度，/>

为与道路有关的待定系数；In the formula:

is the critical threshold of the safety distance; />

is the undetermined coefficient related to speed; />

with />

are the length and width of the vehicle, respectively,

is the undetermined coefficient related to the road;

A3 coordinate transformation

轴方向行驶时，车身会产生一定偏转，车辆场模型也应随之偏转，尤其在换道场景中，公式为：Consider the potential risk that the vehicle will not

When driving in the axial direction, the body will deflect to a certain extent, and the vehicle field model should also deflect accordingly, especially in lane changing scenarios, the formula is:

；

;

为场模型逆时针偏转航向角；It is stipulated that the counterclockwise direction is positive, where:

deflect the heading angle counterclockwise for the field model;

A4 Determining the CAV vehicle field model

，公式为：CAV vehicle field strength described in terms of functions in the form of the Yukawa potential used to describe short-range interactions between nucleons

, the formula is:

；

;

；

;

、/>

均为待定系数；/>

为目标车辆质心所在空间的位置坐标；/>

为目标车辆当前加速度；/>

为目标车辆周围某点到该车辆质心所在空间坐标/>

的夹角；

为原始坐标偏转后的取值；In the formula:

, />

are undetermined coefficients; />

is the position coordinates of the space where the center of mass of the target vehicle is located; />

is the current acceleration of the target vehicle; />

Space coordinates from a point around the target vehicle to the center of mass of the vehicle />

the included angle;

is the value after the original coordinate deflection;

A5 Calibrate the parameters of the CAV vehicle field model and constrain its scope of action

Since the potential field strength in the vehicle field model is limited by the range of action, only a short-range force is applied to the surrounding vehicles within a short distance, and the force changes with the change of vehicle speed, acceleration, distance and other factors. When the target vehicle and the surrounding When the vehicle distance exceeds a certain range, the impact on the field strength of surrounding vehicles is negligible. Using the differential evolution algorithm, the objective function is to minimize the difference between the force range of the CAV vehicle field and the time headway in the strong car-following state, and the performance parameters of the undetermined coefficients in the established CAV vehicle field model are calibrated to make it The field strength distribution is more in line with the real driving state of the vehicle. Taking the headway of strong car-following as the critical action range of the field strength of the CAV vehicle, the field strength value of the CAV vehicle can be obtained, thereby constraining the action range of the CAV vehicle field;

Step B1: Constrain the scope of the HDV vehicle field

The HDV vehicle can maintain a stable state when following the vehicle ahead with a headway of 2.7s. At this time, the field strength of the vehicle is a critical value of 0, namely:

；

;

为标准车头时距；/>

为自车车速；In the formula:

is the standard headway; />

is the vehicle speed;

Step B2: Build the driver's environmental psychological tolerance

Using fuzzy theory, select the typical characteristic factors that reflect the characteristics of the driver: individual experience, environmental visibility, and psychological trust, determine the input eigenvalues of the driver's environmental psychological tolerance evaluation model, and fuzzify the eigenvalues. Through fuzzy rules Logical operation to obtain a fuzzy quantity of driver's environmental psychological tolerance, and then use defuzzification to convert the fuzzy quantity into a precise specific value, which is the driver's environmental psychological tolerance finally calculated by the driver's environmental psychological tolerance evaluation model. Tolerance;

Step B3: Determine the HDV vehicle field model

The CAV vehicle field is a physical quantity whose variable is space-time, so the potential field is essentially a physical field. In physics, due to the law of conservation of momentum, it is believed that momentum exists in the field, and the physical field is real. And the field is a kind of spatial area where each position is affected by force, and the field line is usually used to represent the intensity of the field at a certain position. Therefore, people-vehicle-road in the transportation system all have physical fields due to their own attributes;

However, road users in the traffic system, that is, drivers or pedestrians, are more concerned about obstacles or threats in the direct field of view in the direction of their movement, which directly affect their safety judgments and movement-related behaviors. When applying safety field theory to road users, the interacting vehicle field also creates a "repulsive force" that keeps two interacting human-vehicle combinations at a safe physical gap between each other. But this "action force" that symbolizes risk is not an actual physical force, as it does not obey Newton's laws, such as the equality of action and reaction, but a psychological force, the effect of which is achieved only by the interacting road users behavior manifested. Therefore, different from the physical field, the safety field theory in the context of road users is a psychological field, which only manifests the influence on road user behavior through the psychological effect of risk force;

Thus, the "forces" experienced by drivers and pedestrians are intrinsically linked to their perception. Although under normal circumstances, it would repel interacting road users, there are unwary road users who do not "feel" risky forces and act accordingly even though the situation may be critical. In fact, this failure to act in time according to the forces in the traffic interaction can lead to traffic conflicts and accidents;

Based on the discussion of the above-mentioned psychological force of road users, different from the physical potential field generated by the road environment and CAV vehicles due to their own quality, speed and other attributes, the modeling of the driver's safety field in the present invention fully considers the impact of the driver's personality on driving safety. Based on the differences in judging gender, the road visibility is regarded as an important indicator that affects the driver's psychological state, and the potential field that reflects the driver's psychological response is established as the vehicle field model of the traditional vehicle. The decision-maker of a traditional car is affected by the change of external traffic conditions and produces a psychological force that changes his subjective cognition. The field strength of the traditional vehicle field indicates the degree of coercion the driver has on the conflicting object in the external traffic space. The field strength and the psychological force In the same direction, the greater the degree of external coercion perceived by the driver, the greater the psychological field force received, and the greater the field strength value;

Based on the above discussion, the calculation formula of HDV vehicle field strength is derived as follows:

；

;

为HDV车辆场场强；/>

为与函数极值有关的调整系数； />

为驾驶员环境心理承受度；In the formula:

is the HDV vehicle field strength; />

is the adjustment coefficient related to the extremum of the function; />

The psychological tolerance of the driver's environment;

During the driving process of the vehicle, the road markings and the boundaries on both sides will have a certain restrictive effect on the driving of the vehicle. By restricting the vehicle to keep driving in the middle of the lane line, the continuity of the traffic flow is maintained, and the relatively stable distance between the front and the lateral distance is maintained. Therefore, when the vehicle is driving in the center of the lane, the driving risk is small, and when it deviates close to the road boundary, it will cause a greater risk than crossing the marking line. Assuming that the vehicle is traveling on a two-lane road, it represents the road boundary line on the outside of the lane, which may cause a substantial collision with the vehicle if it is too close;

In order to avoid sudden changes, an exponential function is selected to construct an environmental field with the lateral distance as a variable, and its value is infinite at the boundary to prevent the vehicle from leaving. The formula is:

；

;

表示车道外侧道路边界线，过近可与车辆发生实质性碰撞；/>

为车辆行驶坐标；/>

与/>

分别为道路标线与边界风险场的风险系数；/>

为车道总宽度；/>

为势场收敛系数；In the formula: assuming that the vehicle travels on a two-lane road,

Indicates the road boundary line on the outside of the lane, too close to a substantial collision with the vehicle; />

is the coordinates of the vehicle; />

with />

are the risk coefficients of road markings and boundary risk fields respectively; />

is the total width of the lane; />

is the potential field convergence coefficient;

Aiming at the problem of the change of the number of lanes, the environmental field along the direction of traffic flow is established for the lane, and the repulsive potential of the potential field theory is selected to construct the vehicle boundary environment field, so as to quantitatively characterize the influence of the reduction of the number of lanes on the driving safety of the vehicle during the change of road conditions. The expression as follows:

；

;

为道路标线与边界风险场的风险系数；/>

为车辆/>

到加速车道末端的矢量距离；In the formula:

is the risk coefficient of the road marking and boundary risk field; />

for vehicles />

Vector distance to the end of the acceleration lane;

Draw CAV vehicle field, HDV vehicle field, multi-vehicle superposition vehicle field and environment field through MATLAB;

，车辆加速度/>

，车辆偏航角/>

，车辆质量/>

，车辆位置坐标，以及用差分进化方法标定出的/>

、/>

、/>

、/>

；输出：CAV车辆场场强/>

；Refer to Figure 2-5, CAV vehicle field, input: vehicle speed

, vehicle acceleration />

, vehicle yaw angle />

, vehicle mass/>

, the vehicle position coordinates, and the calibrated /> by the differential evolution method

, />

, />

, />

;Output: CAV vehicle field strength />

;

、驾驶员环境心理承受度

、车辆速度/>

；车辆位置坐标；输出：HDV车辆场场强/>

；See Figure 6, HDV vehicle field, input: Adjustment factor related to the extremum of the function

, The psychological tolerance of the driver's environment

, vehicle speed />

;Vehicle position coordinates; Output: HDV vehicle field strength/>

;

、/>

，两车加速度/>

、/>

，两车偏航角/>

、/>

，两车质量/>

、/>

，两车位置坐标，以及用差分进化方法标定出的/>

、/>

、/>

、/>

；输出：多车叠加后车辆场场强/>

；Refer to Figure 7, multi-vehicle superimposed vehicle field, input: speed of two vehicles

, />

, the acceleration of the two vehicles />

, />

, the yaw angle of the two vehicles />

, />

, mass of two cars/>

, />

, the position coordinates of the two vehicles, and the /> calibrated by the differential evolution method

, />

, />

, />

;Output: vehicle field strength after multi-vehicle superposition/>

;

，道路标线与边界风险场的风险系数/>

、/>

，势场收敛系数/>

；输出：以横向距离为变量的环境场场强

；Refer to Figure 8, the environmental field with the lateral distance as a variable, input: vehicle coordinates, total width of the lane

, the risk coefficient of the road marking and boundary risk field />

, />

, potential field convergence coefficient />

;Output: Ambient field strength as a variable in lateral distance

;

，车辆/>

到加速车道末端的矢量距离/>

；输出：以纵向距离为变量的环境场/>

。Refer to Figure 9, the environmental field with the longitudinal distance as a variable, input: the risk coefficient of the road marking and boundary risk field

, vehicle />

vector distance to end of acceleration lane />

;Output: ambient field with longitudinal distance as variable/>

.

Further, the specific method of said A5 is as follows:

A5.1 Screening the dataset of natural vehicle driving trajectories

A5.2 Preprocessing the dataset

Statistical information about vehicle lane distribution and speed distribution is obtained. However, there are various driving states of vehicles in the process of driving in this area. Therefore, this patent sets the filter conditions of car-following state to process the data. First, ensure that the front and rear vehicles are driving in the same lane; second, set the distance between the front and rear vehicles within the range of 2-150m. When the distance is too small, the default is to eliminate vehicles in queuing state; Time, if the duration is too large or too small, it will be rejected by default as the failure to reach a stable state. By arranging and analyzing the trajectory data of the data set, the distribution interval of each car-following vehicle data is obtained, and the effective car-following data is extracted by analyzing its distribution law;

A5.3 Obtain the time headway under strong car-following state

A5.4 Through the differential evolution algorithm, use the effective car-following data to calibrate the performance parameters of the undetermined coefficients in the CAV vehicle field model; take the headway under the strong car-following state as the critical range of the CAV vehicle field strength, and obtain the CAV The field strength value of the vehicle field constrains the scope of action of the CAV vehicle field.

Further, in said B2, the specific construction method of the driver's environmental psychological tolerance is as follows:

B2.1 Determine the indicators of environmental psychological tolerance

Determine the driver's individual experience, environmental visibility, and psychological trust as indicators of environmental psychological tolerance;

B2.2 Determine the domain of each indicator

The evaluation model of the driver's environmental psychological tolerance is three-input and single-output. The discourse interval of the individual experience degree is set as [0,5], the fuzzy subset is {Low, Medium, High}, and the discourse interval of the environmental visibility is [0 ,3], the fuzzy subset is {Near, Middle, Far}, the psychological trust domain is [-3,3], the fuzzy subset is {negative large NB, negative small NS, zero ZO, positive small PS, positive large PB};

B2.3 Determine the membership function and membership degree of each index

When the fuzzy input is set as the driver's individual experience degree and environmental visibility, the triangular state membership function is used, and when the fuzzy input is the psychological trust degree, the Gaussian shape membership function is selected, and the membership degree of each index is obtained according to the membership function;

B2.4 Establish the driver's environment psychological tolerance evaluation model rule table according to the degree of membership

B2.5 Determine the relationship between the three input indicators from the rule table, and get the tolerance value in the range of [0,1], which is the psychological tolerance of the driver's environment.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention. Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (9)

1. The modeling method for the security field of the heterogeneous traffic epidemic car is characterized by comprising the following steps of:

step A: construction of CAV vehicle field model

A1, quantifying target vehicle attributes;

a2, determining pseudo distance;

the dangerous degree of the target vehicle also depends on the relative positions of the target vehicle and surrounding potential risk vehicles, and the actual space distance of the potential risk vehicles, which are close to the target vehicle at different angles, is corrected by referring to Euclidean distance to obtain pseudo distance;

a3, coordinate transformation is carried out

Consider potentially risky vehicle roll

When the vehicle runs in the axial direction, the vehicle body can deflect to a certain extent,the vehicle field model should also deflect accordingly, especially in lane change scenarios, the formula is:

；

the counterclockwise direction is defined as positive, where:

deflecting the course angle counterclockwise for the field model;

a4 determination of CAV vehicle field model

Describing CAV vehicle field strength as a function of the form of the takawa potential to describe short-range interactions between nuclei

；

A5, calibrating parameters of CAV vehicle field model and restraining action range of the model

Using a differential evolution algorithm, and performing performance parameter calibration on undetermined coefficients in the established CAV vehicle field model by taking a difference value between a force range of a minimized CAV vehicle field and a vehicle head time interval in a strong following state as an objective function, so that field intensity distribution of the undetermined coefficients is more in accordance with a real running state of the vehicle, and taking the strong following vehicle head time interval as a critical action range of field intensity of the CAV vehicle field, so that a field intensity value of the CAV vehicle field can be obtained, and the action range of the CAV vehicle field is restrained;

and (B) step (B): construction of HDV vehicle field model

Step B1: restraining the reach of an HDV vehicle field

The HDV vehicle can keep a stable state when following a car ahead with a 2.7s headway, and the field intensity value of the vehicle is a critical value 0, namely:

；

wherein:

is the standard headway; />

The vehicle speed is the self-vehicle speed;

step B2: construction of driver environmental psychological bearing degree

Adopting a fuzzy theory, selecting typical characteristic factors which reflect the characteristics of a driver: determining the individual experience degree, the environment visibility degree and the psychological trust degree, inputting the characteristic value of the driver environment psychological tolerance assessment model, fuzzifying the characteristic value, obtaining a fuzzy quantity of the driver environment psychological tolerance through logic operation of a fuzzy rule, and converting the fuzzy quantity into an accurate specific numerical value by utilizing defuzzification, namely the driver environment psychological tolerance finally calculated by the driver environment psychological tolerance assessment model;

step B3: determining HDV vehicle field model

；

Wherein:

the field strength is HDV vehicle field strength; />

Is an adjustment coefficient related to the extremum of the function; />

The environmental psychological bearing degree is used for the driver; />

A critical threshold for a safe distance; />

Is a coefficient to be determined in relation to speed; />

Is pseudo distance;

step C: construction of an environmental field model with lateral distance as a variable

An exponential function is selected to construct an environment field taking the transverse distance as a variable, so that the value of the environment field is infinite at the boundary to have a trend of preventing the vehicle from driving away;

step D: construction of an environmental field model with longitudinal distance as a variable

Selecting repulsive potential of potential field theory to construct a vehicle boundary environment field so as to quantitatively represent the influence of reduction of the number of lanes in road condition change on vehicle driving safety, and establishing an environment field along the driving direction of the traffic flow for the lanes;

step E: drawing field intensity diagram of driving safety field

And drawing a CAV vehicle field, an HDV vehicle field, a multi-vehicle superposition vehicle field and an environment field through MATLAB.

2. The heterogeneous transportation epidemic safety field modeling method according to claim 1, wherein in the A1, the target vehicle

Equivalent mass of->

The formula of (2) is:

；

wherein:

for the target vehicle->

Equivalent mass of (a); />

For the target vehicle->

Is the actual mass of (3); />

For the target vehicle->

Is a function of the speed of the machine.

3. The method for modeling security of heterogeneous transportation vehicles according to claim 2, wherein in A2, the pseudo distance is calculated

The formula of (2) is:

；

wherein:

and->

Vehicle length and width, respectively, < >>

Is a predetermined coefficient related to the road.

4. A method for modeling security of a heterogeneous transportation epidemic vehicle according to claim 3, wherein in A4, CAV vehicle field intensity is

The formula is:

；

；

wherein:

、/>

all are undetermined coefficients; />

The position coordinates of the space where the mass center of the target vehicle is located; />

The current acceleration of the target vehicle; />

For the spatial coordinates of a point around the target vehicle to the centroid of the vehicle +.>

Is included in the plane of the first part; />

The value of the original coordinate after deflection is taken.

5. The heterogeneous traffic epidemic car security field modeling method according to claim 4, wherein in the A5, the specific method for calibrating the CAV car field model parameters is as follows:

a5.1 screening of vehicle Natural Driving trajectory data set

A5.2 pretreatment of data set

According to screening conditions of the following state, ensuring that front and rear vehicles run on the same lane, setting the distance between the front and rear vehicles within the range of 2-150m, defaulting to vehicle queuing state rejection when the distance is too small, and defaulting to free flow state rejection when the distance is too large; setting a minimum following time of 10s, and defaulting to a state that a stable state cannot be achieved and rejecting when the duration is too large or too small; acquiring a distribution interval of each following vehicle data by arranging and analyzing track data of a data set, and extracting effective following data by analyzing a distribution rule of the distribution interval;

a5.3 obtaining the headway in the Strong heel-and-heel state

A5.4, calibrating performance parameters of undetermined coefficients in the CAV vehicle field model by utilizing effective following data through a differential evolution algorithm; and taking the headway in a strong following state as a critical action range of the CAV vehicle field intensity to obtain a CAV vehicle field intensity value and restraining the action range of the CAV vehicle field.

6. The heterogeneous traffic epidemic car security field modeling method according to claim 5, wherein the specific construction method of the driver environmental psychological bearing degree in B2 is as follows:

b2.1 determining environmental psychological bearing index

Determining the individual experience degree, the environmental visibility and the psychological trust degree of the driver as environmental psychological bearing degree indexes;

b2.2 determining the discourse of each index

The driver environmental psychological bearing degree evaluation model is three-input single-output, an individual experience degree domain interval is set to be [0,5], a fuzzy subset is { Low, medium, high }, a domain interval of environmental visibility is [0,3], the fuzzy subset is { Near, middle, far }, a psychological trust degree domain is [ -3,3], and the fuzzy subset is { negative big NB, negative small NS, zero ZO, positive small PS, positive big PB };

b2.3 determining the membership function and membership of each index

Setting fuzzy input as individual experience degree of a driver and environment visibility, selecting a triangular form membership function, selecting a Gaussian form membership function when the fuzzy input is psychological trust degree, and obtaining membership of each index according to the membership function;

b2.4 establishing a driver environmental psychological bearing evaluation model rule table according to the membership degree

And B2.5, determining the relation among three input indexes by a rule table to obtain a bearing degree value in the range of [0,1], namely the environmental psychological bearing degree of the driver.

7. The method for modeling security of heterogeneous transportation vehicles according to claim 6, wherein in the step C, the environmental field model formula using the lateral distance as a variable is:

；

wherein: assuming that the vehicle travel road is a two-lane,

indicating a lane outside road boundary line, too close to which a substantial collision with the vehicle can occur; />

The vehicle running coordinates; />

And->

Risk coefficients of the road marking and the boundary risk field are respectively; />

Is the total width of the lane; />

Is the potential field convergence coefficient.

8. The method for modeling security of heterogeneous transportation vehicles according to claim 7, wherein in the step D, the environmental field model formula using the longitudinal distance as a variable is:

；

wherein:

risk coefficients of the road marking and the boundary risk field; />

For vehicle->

Vector distance to the end of the acceleration lane.

9. The method for modeling a security field of a heterogeneous transportation epidemic vehicle according to claim 8, wherein in the step E, the specific process of the field intensity diagram of the security field of the transportation is as follows:

CAV vehicle field, input: vehicle speed

Vehicle acceleration->

Yaw angle>

Vehicle mass->

Vehicle position coordinates, and ++specified by differential evolution method>

、/>

、/>

、/>

The method comprises the steps of carrying out a first treatment on the surface of the And (3) outputting: CAV vehicle field strength>

；

HDV vehicle field, input: adjustment coefficients related to function extremum

Driver mental tolerance->

Speed of vehicle

The method comprises the steps of carrying out a first treatment on the surface of the Vehicle position coordinates; and (3) outputting: HDV vehicle field strength->

；

Multiple vehicle superposition vehicle field, input: speed of two vehicles

、/>

Acceleration of two vehicles->

、/>

Yaw angle +.>

、/>

Two-wheeled vehicle mass->

、/>

Two-position coordinates and differential evolution methodMarked->

、/>

、/>

、/>

The method comprises the steps of carrying out a first treatment on the surface of the And (3) outputting: field intensity of vehicle after multi-vehicle superposition>

；

An environmental field with lateral distance as a variable, input: vehicle coordinates, total lane width

Risk coefficient of road marking and boundary risk field +.>

、/>

Potential field convergence factor->

The method comprises the steps of carrying out a first treatment on the surface of the And (3) outputting: ambient field strength as a function of lateral distance +.>

；

Environmental field with longitudinal distance as variable, input: risk coefficient of road marking and boundary risk field

Vehicle->

Vector distance to the end of the acceleration lane +.>

The method comprises the steps of carrying out a first treatment on the surface of the And (3) outputting: environmental field with longitudinal distance as a function>

。

Images