CN-117010134-B - Expressway tunnel section vehicle following modeling method based on visual characteristics of driver

CN117010134BCN 117010134 BCN117010134 BCN 117010134BCN-117010134-B

Abstract

The invention discloses a highway tunnel section vehicle following modeling method based on visual characteristics of a driver, which comprises the steps of 1, selecting a highway tunnel section, obtaining driving state data and vehicle related parameters to obtain a following state of the vehicle, 2, constructing a following vehicle driver acceleration and deceleration decision behavior model, determining following driving decision behaviors under different driving conditions, 3, constructing a vehicle driver utility optimal function based on safe following distance to obtain a minimum following distance based on the current speed, 4, correcting the vehicle driver utility optimal function based on the perceived speed to obtain a corrected minimum safe following distance, 5, establishing a highway tunnel section vehicle following model based on the visual characteristics of the driver, 6, setting a tunnel section to simulate a traffic scene by changing a leading vehicle running state, and 7, simulating the running state of the following vehicle when t >0 based on a simulated traffic scene following vehicle running according to the vehicle following model.

Inventors

CHI YU
CHEN XIAOYUE
JIN YAN
ZHU JUNYI
TANG HONG
CAI JIANHUA
CAI XIAOMIN
SU WEIJI
LIU GUANLIN
Chen Jiongzhen
YANG SHIYING

Assignees

上海城建城市运营(集团)有限公司
上海城建智慧城市运营管理有限公司

Dates

Publication Date: 20260508
Application Date: 20221221

Claims (9)

1. The highway tunnel section vehicle following modeling method based on the visual characteristics of the driver is characterized by comprising the following steps of: step 1, selecting a highway tunnel section, acquiring driving state data and vehicle related parameters of a vehicle driven by the tunnel section, and calculating to obtain a following state of the vehicle; Step 2, constructing an acceleration and deceleration decision behavior model of a driver of the following vehicle according to the following state of the vehicle, and determining following driving decision behaviors of the driver of the following vehicle under different driving conditions; step 3, constructing a vehicle driver utility optimal function based on the safe following distance, and calculating to obtain a minimum following distance based on the current vehicle speed; Step 4, selecting the acceleration and deceleration decision behavior model of the following vehicle driver based on the perceived speed, correcting the utility optimal function of the vehicle driver, and calculating to obtain a corrected minimum safe following distance; Step 5, establishing a highway tunnel road section vehicle following model based on the visual characteristics of the driver according to the corrected minimum safe following distance; Step 6, setting a highway tunnel section simulation traffic scene by changing the running state of the lead vehicle; and 7, simulating a traffic scene based on the expressway tunnel section, running the following vehicle according to the expressway tunnel section vehicle following model based on the visual characteristics of the driver, and simulating the running state of the following vehicle when the simulation t is more than 0.
2. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: Wherein in step 1, the driving state data is obtained through video statistics or real vehicle tests, The driving state data comprises the speed, the acceleration, the following distance, the road section position and the illumination of the following vehicle, The vehicle-related parameters include a length and a width of the vehicle.
3. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: Wherein in the step 2, the acceleration and deceleration decision behavior model of the following vehicle driver is the acceleration and deceleration decision behavior model of the following vehicle driver on the expressway tunnel section, The concrete construction process of the following vehicle driver acceleration and deceleration decision behavior model is as follows: Step 2-1, the following state of the vehicle is represented by the ratio of the running speed difference between the front vehicle and the rear vehicle to the following distance of the following vehicle, and the following state of the vehicle is represented by the following state: θ=dv/d dv=v N+1 -v N wherein θ is a following vehicle driving state, dv is a vehicle speed difference, v N+1 is a following vehicle speed, v N is a front vehicle speed, and d is a following distance; step 2-2, constructing driver acceleration and deceleration decision models of different driving states from two aspects of following vehicles, namely that the speed of the vehicles is lower than that of the front vehicles and higher than that of the front vehicles, wherein the method comprises the following steps: a=αθ 2 +βθ+γ A=αθ 2 +βθ+γ where a is the acceleration of the following vehicle when the speed of the following vehicle is lower than that of the preceding vehicle, A is the acceleration of the following vehicle when the speed of the following vehicle is higher than that of the preceding vehicle, and alpha, beta and gamma are constants.
4. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: wherein in step 3, the driver finds the front object and makes judgment decision in the driving process of the tunnel section, and the method comprises 3 steps, namely, the front object appears in the driving process of the 1 st step, the driver finds and judges the front object as an obstacle or a vehicle to prepare for coping, the driver makes decision in the 3 rd step, adopts deceleration or braking measures to cope with, so as to ensure driving safety, Estimating the total utility of the following vehicles in the driving process of the tunnel road section from two aspects of time utility and safety utility, constructing the optimal function of the vehicle driver utility based on the safety following distance, calculating the minimum following distance based on the current vehicle speed, The time effect is reflected by the following speed, and the safety effect is reflected by the following distance.
5. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 4, wherein: in the step 3, the specific process of calculating the minimum following distance is as follows; step 3-1, in order to ensure the driving safety, the following vehicle speed should meet the following safety conditions: d min ≤d Wherein d min is the minimum safe following distance under the current following speed, and d is the following distance of the following vehicle; step 3-2, the minimum safe following distance calculation method under the current following speed is as follows: d min ＝L N+1 -L N +β Wherein L N+1 is the braking distance of N+1 vehicles, L N is the braking distance of N vehicles, and beta is the distance between N+1 vehicles and N vehicles after stopping 2 vehicles; Step 3-3, when the following safety distance is determined, considering the most dangerous situation, namely, when the front vehicle is in a uniform deceleration braking state, obtaining the minimum following distance as follows: Wherein, beta 0 is the expected safety distance of a driver, v is the speed before the vehicle is braked, t r is the braking delay time, t a is the time of light adaptation or dark adaptation at the entrance and the exit of a tunnel, A N+1 is the maximum deceleration of an N+1 vehicle, and A N is the maximum deceleration of the N vehicle; Step 3-4, when the vehicle is in a following state, the speed difference between the front vehicle speed and the rear vehicle speed is smaller, so that v N+1 and v N are regarded as two approximately equal values, and meanwhile, the acceleration of the two vehicles is regarded as equal without considering the speed reduction performance difference of the vehicles, so that the minimum safe following distance when the vehicle runs following is obtained: L m ＝v N+1 (t r +t a )+β 0 Where L m is the minimum safe following distance at the current following vehicle speed.
6. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: Wherein, the step 4 specifically comprises the following steps: In step 4-1, in daytime driving, the outer driving section of the tunnel is daytime driving, and the inner driving section of the tunnel is regarded as night condition driving, so that the following speed of the vehicle on the tunnel section is corrected according to the perceived speed function of the daytime and night drivers: Δv o ＝v N+1 -f(v N+1 ) Δv i ＝v N+1 -p(v N+1 ) Wherein Deltav o is the speed sensing deviation of the driver on the outer section of the tunnel, deltav i is the speed sensing deviation of the driver on the inner section of the tunnel, f (v N+1 ) is the speed sensing function of the driver on the outer section of the tunnel, and p (v N+1 ) is the speed sensing function of the driver on the inner section of the tunnel; and 4-2, correcting the following speed of the tunnel section by using the speed deviation to obtain the perceived speed of the driver when the driver walks: V=v N+1 +Δv Wherein V is the theoretical highest speed limit value of the modified expressway extra-long tunnel section, and Deltav is the speed perception deviation value of a driver; Step 4-3, calculating to obtain a corrected minimum following distance: L m ＝V(t r +t a )+β 0 Where L m is the minimum safe following distance at the corrected current following vehicle speed.
7. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: In step 5, from the perspective of local optimization of the following vehicle, considering time utility and safety utility, taking the modified minimum safety following distance function as a decision objective function of the following vehicle driver, and taking acceleration and deceleration measures by the following vehicle driver to modify the running speed and the following distance to an optimal state.
8. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: wherein, in step 6, the running state includes a vehicle speed and a vehicle position.
9. The highway tunnel segment vehicle following modeling method based on the visual characteristics of the driver according to claim 1, wherein: in step 7, it is assumed that the lead vehicle changes the motion state according to a preset situation, and the following vehicle operates according to a highway tunnel road section vehicle following model based on the visual characteristics of the driver, and the following vehicle operation state when t >0 is examined is updated as follows: vehicle speed v n+1 (t+Δt)＝v n+1 (t)+a n+1 (t) ×Δt, n=1,.. Vehicle position: in the formula, Δt is the acceleration adjustment time.

Description

Expressway tunnel section vehicle following modeling method based on visual characteristics of driver Technical Field The invention relates to the field of tunnel vehicle safety, in particular to a highway tunnel section vehicle following modeling method based on visual characteristics of drivers. Background The tunnel is used as an important component in a road system, and because of the special tubular structure, the inside and outside driving environments of the tunnel have large difference, and the accident rate of the tunnel section is lower from the aspect of accident times, but the accident severity is higher, wherein the traffic accident shape is mainly rear-end collision. In the driving accident of the tunnel section, the frequency of occurrence of the entrance and exit of the tunnel is higher, and when a vehicle enters and exits the tunnel portal, the phenomenon of 'visual shock' is generated by a driver due to the severe change of light, so that the driver cannot accurately judge the driving environment in front. Meanwhile, the illumination intensity of the light in the plateau area is stronger than that in the plain area, so that the influence of the change of the illumination intensity on a driver is more serious. Therefore, the following model of the expressway extra-long tunnel based on the visual characteristics of the driver is constructed, the traffic flow stability of the tunnel section is improved, and the optimization of the speed limit value of the tunnel section has stronger practical significance. The following model can be divided into three types, namely a driver stimulus-response model, a safe following distance model and a driver psychological and response model. In the driver stimulus response model, whether the following vehicle accelerates is considered to be determined by the relative speed of the two vehicles and the response of the preceding vehicle. In the safe following distance model, the following speed and the following distance of the rear following vehicle are kept safe in order to avoid rear-end collision accidents caused by sudden braking of the front vehicle. However, the above models lack factors on the visual characteristics of the driver, and only consider single driving environment conditions, so as to consider the influence of the driving environment change on the following driving behavior. Because of the special tubular structure of the expressway tunnel, the driving environments inside and outside the tunnel are greatly different, and the influence of the different driving environments on the driving state and the vision of a driver is not reflected in the model. Disclosure of Invention The invention is made to solve the above problems, and an object of the invention is to provide a highway tunnel section vehicle following modeling method based on visual characteristics of a driver. The invention provides a highway tunnel section vehicle following modeling method based on visual characteristics of a driver, which is characterized by comprising the following steps of 1, selecting a highway tunnel section, obtaining driving state data of a vehicle driven by the tunnel section and relevant parameters of the vehicle, calculating to obtain a following state of the vehicle, 2, constructing a following vehicle driver acceleration and deceleration decision behavior model according to the following state of the vehicle, determining following driving decision behaviors of the vehicle under different driving conditions, 3, constructing a vehicle driver utility optimal function based on a safe following distance according to the following driving decision behaviors, calculating to obtain a minimum following distance based on the current speed, 4, correcting the vehicle driver utility optimal function based on the perception speed by using the following vehicle driver acceleration and deceleration decision behavior model, calculating to obtain the corrected minimum safe following distance, 5, building a highway tunnel vehicle following model based on the visual characteristics of the driver according to the corrected minimum safe following distance, 6, setting a traffic scene by changing the preceding vehicle following running state, and 7, simulating a highway tunnel road traffic scene based on the road following vision simulation scene of the vehicle, and 0. The expressway tunnel section vehicle following modeling method based on the visual characteristics of the driver can be further characterized in that in the step 1, driving state data are obtained through video statistics or a real vehicle test, the driving state data comprise the speed, the acceleration, the following distance, the road section position and the illumination of a following vehicle, and vehicle related parameters comprise the length and the width of the vehicle. The expressway tunnel section vehicle following modeling method based on the visual characteristics of the driver is characterized in th