Transportation Letters

The International Journal of Transportation Research

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/ytrl20

Multi-vehicle anticipation-based driver behavior

models: a synthesis of existing research and future
research directions

Sangram Nirmale, Anshuman Sharma & Abdul Rawoof Pinjari

To cite this article: Sangram Nirmale, Anshuman Sharma & Abdul Rawoof Pinjari (2023): Multi-
vehicle anticipation-based driver behavior models: a synthesis of existing research and future
research directions, Transportation Letters, DOI: 10.1080/19427867.2023.2231212

To link to this article: https://doi.org/10.1080/19427867.2023.2231212

Published online: 11 Jul 2023.

Submit your article to this journal

Article views: 60

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ytrl20
TRANSPORTATION LETTERS
https://doi.org/10.1080/19427867.2023.2231212

Multi-vehicle anticipation-based driver behavior models: a synthesis of existing

research and future research directions
a,b
Sangram Nirmale , Anshuman Sharmab,c and Abdul Rawoof Pinjaria,b
a
Department of Civil Engineering, Indian Institute of Science (IISc), Bengaluru, India; bCentre for Infrastructure, Sustainable Transportation and Urban
Planning (CiStup), Indian Institute of Science (IISc), Bengaluru, India; cCivil Engineering Department, Visvesvaraya National Institute of Technology (VNIT),
Nagpur, India

ABSTRACT ARTICLE HISTORY

Multi-vehicle anticipation (MVA) refers to drivers’ ability to consider stimuli from several vehicles ahead in Received 15 February 2022
their maneuvering decisions, such as longitudinal, lateral, and a combination of longitudinal and lateral Accepted 24 June 2023
movements. This paper provides a comprehensive review of MVA-based driver behavior models developed KEYWORDS
for both homogeneous and heterogeneous disordered (HD) traffic streams. Studies on MVA identify various Multi-vehicle anticipation;
advantages of incorporating MVA in driver behavior models, such as superior numerical and behavioral human factors; driver
soundness, plausible parameter estimates, and model outputs, and improved model realism. In addition, our behavior; homogeneous
findings indicate that MVA-based driver behavior models follow a similar pattern of extending the estab­ traffic conditions,
lished single-leader car-following models, considering vehicles that are directly ahead (in the same lane), and heterogeneous traffic
focussing on a fixed number of vehicles ahead. For HD traffic streams, drivers’ also consider stimuli from conditions
vehicles obliquely placed or on either side. Furthermore, this review discusses issues with the current
modeling approaches and suggests future research directions

Introduction
Driver behavior models are used to describe driver decisions in
various traffic scenarios ranging from free-flow to congested traffic,
city traffic to highway traffic. A few examples of well-known driver
behavior models are Newell’s model (Newell 1961), Gipps’ model
(Gipps 1981), Optimal velocity model (Bando et al. 1995), and
Intelligent driver model (Treiber, Hennecke, and Helbing 2000).
At large, these models are founded on the stimulus-response frame­
work that was first introduced in the 1950s at the General Motors
research laboratories (Chandler, Herman, and Montroll 1958;
Gazis, Herman, and Rothery 1961). According to this framework,
a driver’s response is proportional to the stimulus from a vehicle
ahead. In general, an event or a quantity that evokes a response
from the driver is the stimulus. Researchers have considered various
stimuli such as speed difference between the subject vehicle and the
immediate leader, spacing between the subject vehicle and the
immediate leader, and other sources such as gradient, traffic signals,
lane closures, lane markings, etc. A majority of available driver
behavior models assume that drivers respond to the stimulus
from the immediate lead vehicle only (Brackstone and McDonald
1999; Saifuzzaman and Zheng 2014). However, it has been demon­
strated theoretically and empirically that drivers anticipate down­
stream traffic conditions, consider stimuli from surrounding
vehicles (front and side) and respond accordingly (Hoogendoorn
and Ossen 2006; Nirmale, Pinjari, and Sharma 2021). Therefore,
models that consider stimuli from more than one vehicle to
describe driver behavior offer a higher degree of realism. These
driver behavior models are the focus of this review.
Before moving further, we define some terminologies related to
the anticipation by drivers that are commonly and interchangeably
used in the literature. Existing literature refers to the following
terms in the context of anticipation: multi-anticipative, multi-
vehicle anticipation, spatial anticipation, temporal anticipation,
and multiple sources of stimuli (or information). The terms ‘multi-
anticipative’, ‘multi-vehicle anticipation’, and ‘spatial anticipation’
refer to drivers’ ability to consider stimuli from several vehicles
ahead in their maneuvering decisions. The term ‘temporal antici­
pation’ refers to drivers’ ability to anticipate the traffic situation for
the next few seconds and react accordingly. Whereas the term
‘multiple sources of stimuli’ indicates drivers’ ability to consider
stimuli from lead vehicles as well as from other sources (mentioned
above) in their maneuvering decisions. This review focuses on
models incorporating the multi-vehicle anticipation (MVA) beha­
vior of drivers.
The first question that arises is why do drivers consider stimuli
from several vehicles or anticipate their behavior? Drivers antici­
pate the surrounding vehicles’ behavior to adjust their maneuvering
decisions and perform them safely. In homogeneous traffic streams,
while moving in the longitudinal direction, drivers adjust their
acceleration/deceleration and its extent by observing the dynamics
of vehicles directly ahead. For instance, by noticing the slowing
down of vehicles ahead (second leader or third leader), drivers
proactively regulate their acceleration to avoid sudden jerks.
When it comes to lane-changing, consideration of stimuli from
multiple surrounding vehicles becomes more important because
a lane change is a complex and riskier maneuver and involves
interactions with a greater number of vehicles than car-following.
Further, it is well known that the characteristics of heterogeneous
disordered (HD) traffic streams are substantially different than
those of homogeneous traffic streams in terms of traffic composi­
tion, lane discipline, and overall driver behavior (more on this in
Section 3). One can witness longitudinal, lateral, and a combination
of longitudinal and lateral (also known as two-dimensional) move­
ments in HD traffic streams. For a two-dimensional movement to
be safe, drivers must consider the dynamics of vehicles that are not
only directly ahead but also laterally placed. The anticipation of

CONTACT Sangram Nirmale nirmalek@iisc.ac.in Department of Civil Engineering, Indian Institute of Science (IISc), Bengaluru 560012, India
© 2023 Informa UK Limited, trading as Taylor & Francis Group
2 S. K. NIRMALE ET AL.
vehicular movements becomes far more important in HD traffic
streams because vehicles can cut-in at any time (typical HD stream
characteristic) and may cause a safety-critical situation. Hence, it is
no exaggeration to say that multi-vehicle anticipation is an integral
aspect of drivers’ maneuvering decisions. It makes drivers proactive
and thereby better and safer decision-makers (Sharma et al. 2017).
Studies on MVA enumerate various advantages of incorporating
MVA in driver behavior models, such as superior numerical and
behavioral soundness, plausible parameter estimates, and model
outputs, and an improved model realism (Bexelius 1968;
Hoogendoorn and Ossen 2006; Lenz, Wagner, and Sollacher 1999;
Treiber, Kesting, and Helbing 2006). Despite the importance of
considering MVA, models that incorporate MVA are underex­
plored. Although there are some excellent review studies on driver
behavior models (Asaithambi, Kanagaraj, and Toledo 2016; Azam,
Bhaskar, and Haque 2020; Brackstone and McDonald 1999;
Chakroborty, Maurya, and Vikram 2019; Das and Maurya 2018;
Mahapatra, Maurya, and Chakroborty 2018; Moridpour, Sarvi, and
Rose 2010; Munigety and Mathew 2016; Saifuzzaman and Zheng
2014; Toledo 2007), none of them comprehensively reviews MVA-
based driver behavior models. An in-depth exploration of MVA-
based models will reveal prevalent theories behind such models,
approaches to incorporate MVA behavior, shortcomings of the
available approaches, and a way forward to better mimic the
MVA behavior of drivers.
This paper, hence, attempts to fill this gap by providing a critical
review of MVA-based driver behavior models developed for both
homogeneous and heterogeneous traffic streams. The remainder of
this paper is organized as follows. Section 2 presents a methodology
adopted for collecting the relevant literature. Sections 3 and 4
review models that incorporate MVA by drivers in homogeneous
traffic conditions and HD traffic conditions, respectively. Section 5
provides critical findings from the current review paper. Section 6
synthesizes the gaps with the current approaches to incorporate
MVA behavior and identifies directions for future research. Finally,
Section 7 concludes the paper.
Review framework
This paper adopts a systematic review approach as done by Das and
Maurya (2018), Azam, Bhaskar, and Haque (2020), and Arun et al.
(2021), which includes the following steps: 1) identification of
studies using different inclusion and exclusion criteria, 2) classifica­
tion of identified studies, and 3) evaluation of the identified studies.
The following subsections discuss these three steps.
Identification of studies
The study commenced with a comprehensive search in the peer-
reviewed research literature database such as Web of Science, TRID,
ProQuest, Scopus, Google Scholar, and ResearchRabbit using
appropriate keywords such as ‘multi-vehicle anticipation’, ‘multi-
vehicle anticipation driver behavior models’, ‘multi-vehicle antici­
pation-based models’, ‘multi-stimuli driver behavior’, ‘car-
following’, ‘lane-changing’, ‘homogeneous traffic streams’, and ‘het­
erogeneous, disorderly traffic streams’. These phrases were com­
bined using Boolean operators AND and OR to refine the literature
search further. The search provided a list of relevant literature,
including journal articles, conference proceedings, and theses.
Then, this list of studies was thoroughly examined to avoid missing
any relevant literature. The inclusion criterion was that the models’
mathematical expressions should incorporate the multi-vehicle
anticipation behavior. Other studies that discussed this human
factor at a surface level were removed from the list of studies.
Note that the studies that examine models with multi-vehicle
anticipation behavior also assisted us in understanding the concept
of multi-vehicle anticipation behavior.
Classification of identified studies
A reference management software package is utilized to manage the
large number of references collected for this study and to perform
sorting and categorizing tasks. In addition, the data extraction
procedure included a two-step screening procedure. First, the
downloaded papers are categorized based on their Titles,
Abstracts, and Keywords; then, they are organized based on the
specific topics addressed in each paper. We observe that the major­
ity of MVA-based models are extensions of well-known single-
leader driver behavior models such as the Gazis-Herman-Rothery
(GHR) car-following model, optimal velocity model (OVM), full
velocity difference (FVD) model, and intelligent driver model
(IDM). Hence, we review the MVA-based driver behavior models
under the classifications of extensions of these well-known single-
leader models. The functional forms of these single-leader driver
behavior models are provided in Table 1.
Evaluation of the identified studies
In this step, the identified studies are critically reviewed and sum­
marized in Tables 2 and 3. Specifically, the evaluation focuses on the
strength and limitations of each model, components of the model­
ing framework, input and output variables, and data utilized for
calibration. This systematic review led to a firm grasp of state of the
literature and insights into the multi-vehicle anticipation-based
driving behavior modeling requirements for both homogeneous
and HD traffic conditions.
Models that incorporate multi-vehicle anticipation by
drivers in homogeneous traffic streams
We observe that the majority of MVA-based models are extensions
of well-known single-leader driver behavior models, such as the
Gazis-Herman-Rothery (GHR) car-following model, optimal velo­
city model (OVM), full velocity difference (FVD) model, and intel­
ligent driver model (IDM). Hence, we review the MVA-based
driver behavior models under the classifications of extensions of
these well-known single-leader models. The functional forms of
these single-leader driver behavior models are provided in Table 1.
Extensions of GHR car-following model
Bexelius (1968) and Gazis, Herman, and Potts (1959) were probably
the first to extend the well-known GHR car-following model to
consider stimuli from multiple leaders. Equation (1) provides the
model formulation:
X
N
ðjÞ
ai ðt þ Ti Þ ¼ αj ΔVi ðtÞ (1)
j¼1
where, ai ðt þ Ti Þ is the acceleration of the subject vehicle i at time
t þ Ti ; Ti represents reaction time; N is the number of lead
ðjÞ
vehicles considered; ΔVi ðtÞ represents the speed � difference
between the speed of the jth lead vehicle Vj ðtÞ and speed of
the subject vehicle ðVi ðtÞÞ, and αj is sensitivity coefficient (or
ðjÞ
weight) to ΔVi ðtÞ. As can be observed from Equation (1), the
MVA is incorporated by considering cumulative weighted stimuli
from N vehicles ahead. Setting N ¼ 1 gives the original GHR car-
 TRANSPORTATION LETTERS 3

Table 1. Single leader driver behavior models.

Model Formulation
GHR linear car-following model (Chandler, Herman, and Montroll 1958; Gazis, Herman, and Potts 1959) ai ðt þ Ti Þ ¼
ð1Þ
αj ΔVi ðtÞ
GHR non-linear car-following model (Gazis, Herman, and Rothery 1961) ΔVi ðtÞ
ð1Þ
ai ðt þ Ti Þ ¼ αðVi ðt þ Ti ÞÞβ ð1Þ γ
ðΔXi ðtÞÞ
� �
Helly’s model (Helly 1959) ð1Þ ð1Þ
ai ðt þ Ti Þ ¼ α1 ΔVi ðtÞ þ α2 ΔXi ðtÞ ΔXi
ðdesÞ
ðt þ Ti Þ ,
where,
ðdesÞ
ΔXi ðt þ Ti Þ ¼ β0 þ β1 Vi ðtÞ þ β2 ai ðtÞ
h � � i
Optimal velocity model (OVM) (Bando et al. 1995) ð1Þ
ai ðtÞ ¼ αj VOV ΔXi ðtÞ Vi ðtÞ ,
where,
� � � �
ð1Þ ð1Þ
VOV ΔXi ðtÞ ¼ tanh ΔXi ðtÞ 2 tanhð2Þ
h � � i � �
Full velocity difference (FVD) model (Jiang, Wu, and Zhu 2001) ð1Þ
ai ðtÞ ¼ a0 VOV Δ Xi ðtÞ
ð1Þ
Vi ðtÞ þ α 1 Δ Vi ðtÞ
where,
� � � � � �
ð1Þ ð1Þ
VOV ΔXi ðtÞ ¼ V1 þ V2 C1 tanh ΔXi ðtÞ Li C2
" � � � �#
Intelligent Driver Model (IDM) β
ΔXi
ð1;desÞ
ðtÞ
2
Vi ðtÞ
(Treiber, Kesting, and Helbing 2006) ai ðtÞ ¼ amax
i 1 ðdesÞ ð1Þ ,
Vi ðtÞ ΔXi ðtÞ

where, rffiffiffiffiffiffiffiffiffiffiffi ð1Þ

ð1;desÞ Vi ðtÞ ðdesÞ Vi ðtÞΔVi ðtÞ
ΔXi ðtÞ ¼ C0 þ C1 ðdesÞ þ Vi ðtÞTi pffiffimax
ffiffiffiffiffifficonf
ffiffiffiffi
Vi ðtÞ 2 ai bi

Note – Refer to the Appendix A for details on the notation used in the above table.

following model with a single lead vehicle. An advantage of this decisions (acceleration or deceleration). To fill these gaps,
model is its simplicity. However, the model presented in Equation Hoogendoorn, Ossen, and Schreuder (2006) proposed
(1) suffers from the same limitations as the original GHR model. a generalized version of Helly’s (1959) model by including
These are: (i) inter-driver heterogeneity is not captured since the a distance-dependent factor in it as provided below:
model assumes the identical reaction time values for all drivers;
and (ii) it assumes that the drivers’ maneuvering actions are only
N1
X N2
X � �
ðjÞ ðjÞ ðdesÞ
ai ðt þ Ti Þ ¼ αj ΔVi ðtÞ þ βj ΔXi ðtÞ ΔXi ðt þ Ti Þ
dependent on relative speed. Further, the earlier studies did not
j¼1 j¼1
provide empirical evidence for the presence of MVA behavior
(Saifuzzaman and Zheng 2014). (4)
To empirically investigate MVA-based models, Hoogendoorn and ðjÞ
where, ΔXi ðtÞ represents the distance between the lead vehicle j and
Ossen (2006), Hoogendoorn, Ossen, and Schreuder (2006) and ðdesÞ
Zhang (2014) analyzed vehicle trajectory data using regression ana­ subject vehicle i at time t; ΔXi ðt þ Ti Þ is desired distance between
lysis. Hoogendoorn and Ossen (2006) utilized the multivariate linear the subject vehicle i and its jth leader at time ðt þ Ti Þ. In Equation (4),
regression analysis to estimate parameters of Bexelius’s multi- the additive expression brings in relative speed and spacing as different
anticipatory car-following model (Equation (1)). They empirically sources of stimuli. Hoogendoorn, Ossen, and Schreuder (2006)
demonstrated that drivers not only consider the vehicle directly employed the maximum likelihood estimation approach to calibrate
ahead but also the second leader. Furthermore, they concluded that the sensitivities (aka, coefficients) of the stimuli from different vehicles,
the degree of driver’s reaction to the second leader is dependent on and used a likelihood ratio test to evaluate the benefits of accounting for
both the types of the following and followed vehicles. Drivers follow­ MVA. Their empirical results suggested that incorporating the MVA
ing a truck, for example, had a weaker reaction to the second leader effect improved the extent to which the models could explain observed
on average than those behind a car, perhaps because drivers cannot driver behavior.
easily sight past a truck. Similarly, truck drivers have a significant Following Hoogendoorn, Ossen, and Schreuder (2006), Zhang
reaction to the second leader, maybe because truck drivers’ elevated (2014) provided more insights on the linear-type MVA driver beha­
vantage point offers a greater sight distance. vior model. Specifically, Zhang (2014) proposed a modified general­
Next, Hoogendoorn, Ossen, and Schreuder (2006) proposed the ized linear MVA car-following model (Equation (5)) and addressed
following two modifications to the Bexilus’ MVA model. two issues – multicollinearity between explanatory variables and the
Bexelius Type 2 model: serial correlation of time series data.
n o N1
ð1Þ ð2Þ ðNÞ
ai ðt þ Ti Þ ¼ α min ΔVi ðtÞ; ΔVi ðtÞ; . . . ; ΔVi ðtÞ (2) X ðjÞ
ai ðt þ Ti Þ ¼ α0 þ αj ΔVi ðtÞ
j¼1
Bexelius Type 3 model: N2
X � �
ðjÞ ðdesÞ
n o þ βj ΔXi ðtÞ ΔXi ðt þ Ti Þ þ γVi ðtÞ þ εit
ð1Þ ð2Þ ðNÞ
ai ðt þ Ti Þ ¼ min α1 ΔVi ðtÞ; α2 ΔVi ðtÞ; . . . ; αN ΔVi ðtÞ (3) j¼1

These models assume that the driver may only respond to the
first, second, or third leader based on the speed difference
between their vehicle and the lead vehicles. However, these
models do not consider relative spacing as a stimulus and
ignore behavioral differences between different maneuvering
(5)
This study empirically demonstrated that the traffic congestion
level (extremely congested and less congested) affects driver’s reac­
tions to different stimuli with respect to different lead vehicles. For
example, the study’s estimation results depicted that drivers in
 4

Table 2. Summary on models that incorporate multi-vehicle anticipation by drivers in homogeneous traffic streams.
Model category Author(s) Proposed model Data utilized and region Strengths Weakness
Extensions of GHR Car- Bexelius (1968) and Gazis, N
P – ● Simple model is proposed ● The model is too simplis­
ðjÞ
ai ðt þ Ti Þ ¼ αj ΔVi ðtÞ ● Special case of this model
Following Model Herman, and Potts (1959) j¼1
tic to describe actual traf­
gives the GHR model fic phenomena
● Stability conditions are accurately.
derived analytically ● Inter-driver heterogeneity
is not captured since the
model assumes identical
S. K. NIRMALE ET AL.

reaction time values for

all drivers
● Model assumes that the
drivers’ maneuvering
actions depend only on
relative speed
● Empirical evidence for the
presence of MVA beha­
vior is not shown
Hoogendoorn, Ossen, and Bexelius Type 2 model: ● Provide empirical evi­ ● Bexelius Type 2 model
n o • Vehicular trajectory
Schreuder (2006) ð1Þ ð2Þ ðNÞ dence that drivers con­ and Bexelius Type 3
ai ðt þ Ti Þ ¼ α min ΔVi ðtÞ; ΔVi ðtÞ; . . . ; ΔVi ðtÞ data collected at the
three-lane A15 sider not only the first model do not consider
Bexelius Type 3 model:
n o motorway to the leader but also the second relative spacing as
ð1Þ ð2Þ ðNÞ
ai ðt þ Ti Þ ¼ min α1 ΔVi ðtÞ; α2 ΔVi ðtÞ; . . . ; αN ΔVi ðtÞ
South of the Dutch leader a stimulus
Generalized version of Helly’s model city of Rotterdam,
● The issue of serial correla­ ● Models ignore behavioral
PN1 N2
P � � tion and its implications is differences between dif­
ðjÞ ðjÞ ðdesÞ Netherlands
ai ðt þ Ti Þ ¼ αj ΔVi ðtÞ þ βj ΔXi ðtÞ ΔXi ðt þ Ti Þ addressed through the ferent maneuvering deci­
j¼1 j¼1 Vehicular trajectory data
collected A2 motor­ Durbin-Watson test sions (acceleration or
way near the Dutch
● Well-known multi- deceleration)
city of Utrecht, anticipative driver beha­
Netherlands vior models are compared
empirically
● Well-known multi-
anticipative driver beha­
vior models are compared
empirically
● Consideration of driver
heterogeneity and its
implications is discussed
Zhang (2014) N1
P N2
P � � ● Empirically investigates ● Models ignore behavioral
ðjÞ ðjÞ ðdesÞ
ai ðt þ Ti Þ ¼ α0 þ αj ΔVi ðtÞ þ βj ΔXi ðtÞ ΔXi ðt þ Ti Þ þ γVi ðtÞ þ εit • NGSIM data of vehicle multi-anticipative car- differences between dif­
j¼1 j¼1 trajectories on
a segment of U.S. following behavior using ferent maneuvering deci­
Highway 101, Los a multiple linear regres­ sions (acceleration or
Angeles, United States sion approach is provided deceleration)
● Multicollinearity between
explanatory variables and
the serial correlation in
time series data are
addressed
● Results show that a driver
in extremely congested
traffic conditions reacts to
stimuli from the
first, second, and even the
third leader
(Continued)
Table 2. (Continued).
Model category
Extensions of Models Based Lenz, Wagner, and Sollacher
on Optimal Velocity Model (1999)
Extensions of full velocity
difference (FVD) model
Author(s) Proposed model Data utilized and region Strengths Weakness
N
� � ðjÞ � �
P ΔXi ðtÞ – ● Special case of this model ● The same functional form
ai ðtÞ ¼ αj VOV j Vi ðtÞ
j¼1
gives the optimal velocity for VOV ð:Þ that was used
model in the original optimal
● It is shown that considera­ velocity model is
tion of more than one employed
lead vehicle leads to sta­ ● Empirical evidence for the
bilization of the dynami­ presence of MVA beha­
cal behavior vior is not shown
● Linear stability analysis is
performed
h � � i
Hasebe, Nakayama, and ð1Þ ð2Þ ðJÞ – ● The study demonstrates ● The same functional form
ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; VðΔXi ðtÞ; . . . ; VðΔXi ðtÞ Vi ðtÞ
Sugiyama (2003) that the proposed model for VOV ð:Þ that was used
provides greater stability in the original optimal
than the original optimal velocity model is
velocity model employed
!
Ge et al. (2004) N
P – ● The study shows that ● The simulation results
ðjÞ
ai ðt þ Ti Þ ¼ VOV αj ΔXi ðtÞ anticipating the behavior from the study confirmed
j¼1
of more vehicles ahead of that only the information
the subject vehicle results of three cars ahead of the
in the stabilization of subject vehicle is enough
traffic systems for cooperative driving.
However, empirical evi­
dence is missing from this
study.
Wilson et al. (2004) Model A: – ● Issue of unrealistic predic­ ● Formal analysis is missing
( )
PN � � tion of acceleration values regarding which formula­
ðjÞ
ai ðtÞ ¼ α0 αj VOV ΔXi ðtÞ Vi ðtÞ by the original optimal tion (Model A and Model
j¼1
velocity model is B) is better
Model B:
( � ðjÞ � ) addressed
PN
ai ðtÞ ¼ α0 αj VOV
ΔXi ðtÞ
Vi ðtÞ
● Weighted sum of the ori­
j
j¼1 ginal optimal velocity
function is considered
( )
Chen, Shi, and Hu (2012) PN ● Reaction time delay is ● The study majorly focuses
ðjÞ
ai ðt þ Ti Þ ¼ α0 αj VOV ðΔXi ðtÞÞ Vi ðtÞ incorporated in the multi- on stability analysis and
j¼1
vehicle anticipation driver does not attempt to esti­
behavior model mate the model
● The study shows that
multiple look-ahead con­
siderations could mitigate
the negative impact due
to reaction time delay on
traffic flow stability
h m
i P � � ● Velocity differences of ● Empirical evidence for
Wang, Gao, and Zhao (2006) ðdesÞ ðjÞ ðjÞ –
ai ðtÞ ¼ a0 Vi ðΔXi ðtÞÞ Vi ðtÞ þ αj ΔVi ðtÞ
j¼1 multiple vehicles are con­ MVA behavior is not
sidered to enhance the provided
stability of traffic flow as
compared to the FVD
model
TRANSPORTATION LETTERS
(Continued)
5

Table 2. (Continued).
Model category
Li and Liu (2006)
Yu, Shi, and Zhou (2008)
Peng and Sun (2010)
Jin, Xu, and Gao (2011)
Li et al. (2011)
Extensions of Intelligent Treiber, Kesting, and Helbing
Driver Model (IDM)
6
Author(s) Proposed model Data utilized and region Strengths Weakness
!
N
P � – ● The study confirms that ● Model was not estimated
ð1Þ ðjÞ 1 j
ai ðt þ Ti Þ ¼ VOV ΔXi ðtÞ; αj ΔVi ðtÞ ; where αj ¼ 5 incorporating the relative using any data
j¼1
speed as a stimulus can
stabilize the traffic flow
just as in the forward-
S. K. NIRMALE ET AL.
looking optimal velocity
model
h � � i h i
ðjÞ ðjþ1Þ ðjþm 1Þ ð1Þ – ● The study demonstrates ● The model is not cali­
ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; ΔXi ðtÞ; . . . ; ΔXi ðtÞ Vi ðtÞ þ α1 ΔVi ðtÞ
that traffic jams are sup­ brated using real vehicu­
pressed more efficiently lar trajectory data
by considering the head­ ● Reaction time is ignored
way of more lead vehicles in this model
ahead of the subject
vehicle and relative speed
with respect to the first
lead vehicle
h i
ðjÞ ðjþ1Þ ðjþN 1Þ – ● Numerical simulation ● Empirical evidence for the
ai ðtÞ ¼ α0 VOV ðΔXi ðtÞ; ΔXi ðtÞ; . . . ; ΔXi ðtÞÞ Vi ðtÞ þ
� � results show that traffic presence of MVA beha­
ðjÞ ðjþ1Þ ðjþN 1Þ
β0 G ΔVi ðtÞ; ΔVi ðtÞ; . . . ; ΔVi ðtÞ jams are reduced effec­ vior is not shown
tively when more lead ● The model does not
vehicles are considered account for heterogene­
ity in driving behavior
� � � �
ð1Þ ð2Þ ð1Þ ð2Þ – ● The study demonstrates ● The model considers sti­
ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ þ α1 ΔVi ðtÞ þ α2 ΔVi ðtÞ
that considering the gen­ muli from only two lead
eralized optimal velocity vehicles
function and velocity dif­ ● Reaction time is ignored
ferences with respect to in this model
multiple vehicles stabi­ ● The same functional form
lizes traffic flow and sup­ for VOV ð:Þ that was used
presses traffic jams in the original optimal
velocity model is used
" ! # " # " #
N
P PN � � N
P � � – ● In addition to space head­ ● Reaction time is ignored
ðjÞ ðjÞ ðjÞ
ai ðtÞ ¼ λ0 VOV αj ΔXi ðtÞ Vi ðtÞ þ λ1 βj ΔVi ðtÞ þ λ2 γj Δai ðtÞ ways and velocity differ­ in this model
j¼1 j¼1 j¼1
ences, acceleration ● The functional form for αj
differences are also incor­ is the same as that of
porated into the driver Peng and Sun (2010)
behavior model
● The study illustrates that
the proposed model had
a greater stable region
than that of the full velo­
city difference model
0 12
" � �β # – ● Looking several vehicles ● Reaction time is not con­
N
P
(2006) Vi ðtÞ B ΔXiðj;desÞ ðtÞ C ahead (spatial anticipa­ sidered in this model
ai ðtÞ ¼ amax
i 1 ðdesÞ amax
i @PN A
Vi ðtÞ j¼1 ðjÞ tion) is considered in
ΔXi ðtÞ
j¼1
Intelligent Driver Model
● String stability analysis is
performed
(Continued)
Table 2. (Continued).
Model category
Chen et al. (2010)
Extension of Collision- Eissfeldt and Wagner (2003)
Avoidance Models
Extensions of Piece-Wise Farhi et al. (2012)
Linear Car-Following
Model
Author(s) Proposed model Data utilized and region Strengths Weakness
" � �β # � ðjamÞ �2
N
P – ● An issue of the intelligent ● The model does not
Vi ðtÞ ΔXi ðtÞþViequ ðtÞTi
ai ðt þ Ti Þ ¼ amax
i 1 ðdesÞ amax
i αj ΔXequ driver model that is dif­ account for heterogene­
Vi ðtÞ j¼1
ferent desired headway at ity in driving behavior
the same uniform space ● Calibration of the model
gap is addressed is not done
● This study also provides
a model which is
a combination of the
optimal velocity model
and full velocity differ­
ence model
● Linear stabilities of differ­
ent multi-anticipative
intelligent driver models
are performed
Vpred ¼ maxfVdes εa; 0g – ● Role of anticipation is ● Fixed set of parameters is
n � � o
ð2�Þ explored through simula­ used in simulation
Vdes ¼ min V1 þ a; V2;safe V2 ; ΔXi ; Vmax
� � q ffiffiffi ffiffiffiffi ffiffiffiffi ffiffiffiffi ffiffiffi ffiffiffiffi ffiffiffi ffiffiffiffiffi ffiffiffiffiffiffiffiffiffiffiffi tion and analytical instead of calibrating
ð2�Þ ð2�Þ
Vsafe V2 ; ΔXi ¼ bTi þ b2 Ti2 þ V22 þ 2bΔXi calculation them using real trajectory
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi�
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi�ffiffi data
� �
Vi;safe ¼ 2 þ 2b ΔX ð1Þ þ V
bTi þ b2 Ti2 þ Vpre i pred Ti min Vpred Ti ; ΔXconstant
� � ðjÞ � �
1 ΔXi ðtÞ ● Min-plus algebra is used ● The model is tested only
Xi ðt þ 1Þ ¼ Xi ðtÞ þ min ð1 þ λÞj min min αuw j þ βuw • NGSIM data of vehicle
1�j�N u2U w2W to model driver behavior for single-driver trajectory
trajectories on
a segment of U.S.
● Car velocities are mod­ data
Highway 101, Los eled as a function of the ● The model does not
Angeles, United States inter-vehicular distances account for heterogene­
● Minimum form is used ity in driving behavior
rather than additive form
TRANSPORTATION LETTERS
7
8 S. K. NIRMALE ET AL.

extremely congested traffic conditions would react to relative Wilson et al. (2004) proposed multiple look-ahead models that
speeds with respect to the first, second, and even the third leader. consider information from multiple leaders as provided below.
Model A:
( )
X
N � �
ðjÞ
Extensions of models based on optimal velocity model ai ðtÞ ¼ α0 αj VOV ΔXi ðtÞ Vi ðtÞ (11)
j¼1
Extensions of optimal velocity model
The optimal velocity model (OVM) was also widely utilized to Model B:
mimic the MVA behavior of drivers. For example, Lenz, Wagner, ( ! )
X
N ðjÞ
and Sollacher (1999) extended the OVM to include multiple vehicle ΔXi ðtÞ
ai ðtÞ ¼ α0 αj VOV Vi ðtÞ (12)
responses. Following Bexelius (1968), Lenz, Wagner, and Sollacher j
j¼1
(1999) assumed that drivers react to the dynamics of their leading
P
vehicle and an arbitrary number of vehicles ahead with a sensitivity In Model A, αj > 0 and Nj¼1 αj ¼ 1 so that uniform flow solutions
αj . The mathematical formulation is provided below: are the same as in the standard OVM, thus, if the space gap between
" ! # the subject vehicle i and the lead vehicle j is small, the subject
XN ðjÞ
ΔXi ðtÞ vehicle i will have a lower optimal velocity than the standard
ai ðtÞ ¼ αj VOV Vi ðtÞ (6)
j¼1
j OVM. Hence, the subject vehicle i will tend to brake earlier when
approaching congested traffic. Model B is taken from Lenz,
Here, N ¼ 1 leads to the original optimal velocity model. They Wagner, and Sollacher (1999) where the optimal velocity of the
borrowed the following functional form from Bando et al. (1998)1 subject driver i is given by a weighted sum of the original optimal
for VOV ð:Þ. velocity function, evaluated at a distance to each lead vehicle ahead.
� � � �
ð1Þ ð1Þ The study exhibited that the proposed model did not predict
VOV ΔXi ðtÞ ¼ tanh Δ Xi ðtÞ C
unrealistic values of acceleration. Note that we observe two types
tanhðCÞ where C is aconstant (7) of representations when MVA behavior is considered in the OVM.
In Equations (11) and (12), the sensitivity coefficient is directly
The study demonstrated increased stability with the MVA-based
multiplied to the optimal velocity function, whereas in Equations
model. Same as in the case of the GHR model, the MVA is incor­
(6) and (8), the sensitivity coefficient is multiplied by the difference
porated by summing up the weighted stimuli from vehicles ahead.
between optimal velocity function and velocity of the subject vehi­
However, a functional form is assumed to compute the sensitivity
cle. Note that formal analysis is missing in the existing literature
coefficient αj .
regarding which formulation is better. We believe that the latter is
Notably, a few studies have extended Lenz’s model by consider­
more behaviorally sound since it considers driver’s sensitivity to the
ing reaction time and desired distance (Chen et al. 2016;
stimuli; i.e. the difference between the optimal velocity and the
Hoogendoorn, Ossen, and Schreuder 2006; Hu, Ma, and Chen
velocity of the subject vehicle rather than the optimal velocity only.
2014). For instance, Hoogendoorn, Ossen, and Schreuder (2006)
Furthermore, Chen, Shi, and Hu (2012) incorporated driver’s
integrated the reaction time as per Equation (8).
" ! # reaction time in Wilson et al. (2004)’s model and proposed
XN ðjÞ
ΔXi ðtÞ a multiple look-ahead model with driver reaction delay as below:
ai ðt þ Ti Þ ¼ αj VOV Vi ðtÞ (8) ( )
j¼1
j X
N
ðjÞ
ai ðt þ Ti Þ ¼ α0 αj VOV ðΔXi ðtÞÞ Vi ðtÞ (13)
Later, Ge et al. (2004) proposed an extended car-following model by j¼1
incorporating headways of arbitrary numbers of lead vehicles in
optimal velocity function itself. The functional form of the model is Following Ge et al. (2004), the following functional form was
given below: employed by Chen, Shi, and Hu (2012) for αj .
! X N �N 1
X N j for j�N
N
ai ðt þ Ti Þ ¼ VOV
ðjÞ
αj ΔXi ðtÞ (9) αj ¼ 1; αj ¼ 1 (14)
j¼1 N j 1 for j ¼ N:
j¼1
ðjÞ
where αj is a weighted function of ΔXi ðtÞ with the following The study utilized the following optimal velocity function from the
properties: Bando et al. (1995) study.
� � h � � i
ðJÞ ðjÞ
(1) αj decreases monotonically as j increases, i.e. α1 > α2 indicat­ VOV ΔXi ðtÞ ¼ 0:5Vmax tanh ΔXi ðtÞ ΔXc þ tanhðΔXc Þ
ing that the influence of the vehicle ahead of the subject (15)
vehicle reduces gradually as the distance between the subject
vehicle and lead vehicle increases. where, Vmax is the maximum velocity and ΔXc is the safety distance.
(2) αj takes the following functional form: Hasebe, Nakayama, and Sugiyama (2003) also extended the
�6 OVM to incorporate MVA behavior and called this model
X N
for j�N a forward-looking optimal velocity model (FLOVM). Their formu­
αj ¼ 1; α1 ¼ 1 "j ¼ 1; αj ¼ 7j1 (10)
7 j 1 for j ¼ N: lation is given below:
j¼1
h � � i
ð1Þ ð2Þ ðJÞ
The authors performed a simulation-assisted stability analysis and ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; VðΔXi ðtÞ; . . . ; VðΔXi ðtÞ Vi ðtÞ
found that MVA further stabilizes the traffic flow compared to the
(16)
single leader model. Additionally, the simulation results from the
study confirmed that only the information of three cars ahead of the The authors employed the optimal velocity function of Bando et al.
subject vehicle is enough for cooperative driving. However, empiri­ (1995) and demonstrated that the FLOVM provides greater stability
cal evidence is missing from this study. than the original OVM.
 TRANSPORTATION LETTERS 9

Extensions of full velocity difference (FVD) model

The FVD model (which is also based on the OVM) was extended by
where;� � h � � i
Wang, Gao, and Zhao (2006) using the velocity differences of
ðjÞ ðjÞ
multiple vehicles as provided in Equation (17) and called it V � ΔXi ðtÞ ¼ 0:5Vmax tanh ΔXi ðtÞ ΔXc þ tanhðΔXc Þ
a multiple velocity difference (MVD) model.
(23)
h i X
m � �
ðdesÞ ðjÞ ðjÞ
ai ðtÞ ¼ a0 Vi ðΔXi ðtÞÞ Vi ðtÞ þ αj ΔVi ðtÞ (17) � �
ðjÞ ðjþ1Þ ðjþN 1Þ ðjÞ
j¼1 G ΔVi ðtÞ; ΔVi ðtÞ; . . . ; ΔVi ðtÞ ¼ β1 ΔVi ðtÞ
ðjþ1Þ ðjþN 1Þ
The MVD model offered more stability and a better suppres­ þβ2 ΔVi ðtÞ þ . . . þ βN ΔVi ðtÞ (24)
sion of traffic jams than the FVD model. The same strategy of To reflect that the influence of a lead vehicle on the subject vehicle
cumulative weighted stimuli is adopted in FVD-based exten­ reduces with its distance from the subject vehicle, the authors
sions too. proposed the following functional form for αj and βj
Next, by introducing relative speed of multiple lead vehicles �N 1
ahead of the subject vehicle in the FVD Model, Li and Liu (2006) X N
1
αj ¼ 1; αj ¼ N j for j�N; βj ¼ J 1 (25)
proposed a forward-looking relative velocity (FLRV) model as 1
j 1 for j ¼ N; N
j¼1 N
given below:
! � �j where, αj and βj are the weighting values representing intensity that
XN
1
ð1Þ
ai ðt þ Ti Þ ¼ VOV ΔXi ðtÞ;
ðjÞ
αj ΔVi ðtÞ ; where αj ¼ driver reacts to its jth lead vehicle. Note that Ge et al. (2004) and
j¼1
5 Chen, Shi, and Hu (2012) implemented a similar functional form
(18) for αj in their study. Numerical simulation results by Peng and Sun
(2010) showed that traffic jams are reduced effectively when more
The modified optimal velocity function is given below: lead vehicles are considered.
! Further, Jin, Xu, and Gao (2011) proposed an extended car-
XN � �
ð1Þ
VOV ΔXi ðtÞ;
ðjÞ ð1Þ
αj ΔVi ðtÞ ¼ tanh ΔXi ðtÞ 4 tanhð4Þ following model that is based on original OVM and FVD models
j¼1 by adopting multiple velocity differences, as below:
! � � � �
X N ð1Þ ð2Þ ð1Þ
þλ
ðjÞ
αj ΔVi ðtÞ ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ þ α1 ΔVi ðtÞ
j¼1 ð2Þ
þ α2 ΔVi ðtÞ (26)
(19)
The authors used the following generalized optimal velocity
where, λ is a constant that is independent of time, velocity, and function.
position. In addition, this study confirmed that incorporating the � �
ð1Þ ð2Þ ð1Þ ð2Þ
relative speed as a stimulus can stabilize the traffic flow just as in the VOV ΔXi ðtÞ; ΔXi ðtÞ ¼ ð1 pÞΔXi ðtÞ þ pΔXi ðtÞ; where;
FLOVM. 1
Later, Yu, Shi, and Zhou (2008) proposed an extended model 0�p� (27)
2
(Equation (20)) that includes a special case of the original OVM and ð1Þ ð2Þ
FVD model. Note that ΔXi ðtÞ is weighted more as compared to ΔXi ðtÞ to
h � � i reflect that the first leader’s influence is greater than the second
ðjÞ ðjþ1Þ
ai ðtÞ ¼ α0 VOV ΔXi ðtÞ; ΔXi
ðjþm
ðtÞ; . . . ; ΔXi
1Þ
ðtÞ Vi ðtÞ leader’s influence. Further, the study demonstrated that con­
h i sidering the generalized optimal velocity function and velocity
ð1Þ
þ α1 ΔVi ðtÞ (20)
differences with respect to multiple vehicles stabilizes traffic
The authors employed the following optimal velocity function: flow and suppress traffic jams.
" !
Previously discussed studies used only the space headways
� � X
m
and relative velocities with respect to multiple lead vehicles to
ðjÞ ðjþ1Þ ðjþm 1Þ ðjÞ
VOV ΔXi ðtÞ; ΔXi ðtÞ; . . . ; ΔXi ðtÞ ¼ 0:5Vmax tanh αj ΔXi ðtÞ ΔXc
j¼1 analyze driver behavior. However, Li et al. (2011) argued that in
þ tanhðΔXc Þ� (21) addition to space headways and velocity differences, accelera­
tion differences could also influence driver behavior. Therefore,
where, ΔXc is the constant safe distance. This study demonstrated
they proposed the multiple headway, velocity, and acceleration
that traffic jams are suppressed more efficiently by considering the
difference model that considered all three types of stimuli. The
headway of more lead vehicles ahead of the subject vehicle and
functional formulation is given below.
relative speed with respect to the first lead vehicle. " ! #
Similarly, Peng and Sun (2010) also modified the FVD model to X N
ðjÞ
propose the following MVA model. ai ðtÞ ¼ λ0 VOV αj ΔXi ðtÞ Vi ðtÞ
h i " j¼1
# " #
ðjÞ ðjþ1Þ ðjþN 1Þ � � � �
ai ðtÞ ¼ α0 VOV ðΔXi ðtÞ; ΔXi ðtÞ; . . . ; ΔXi ðtÞÞ Vi ðtÞ X
N
ðjÞ
XN
ðjÞ
� � þ λ1 βj ΔVi ðtÞ þ λ2 γj Δai ðtÞ (28)
ðjÞ ðjþ1Þ ðjþN 1Þ
þβ0 G ΔVi ðtÞ; ΔVi ðtÞ; . . . ; ΔVi ðtÞ (22) j¼1 j¼1

where, VOV ð:Þ is the optimal velocity function and Gð:Þ is assumed where, λ0 > 0, and λ1 ; λ2 2 ½0; 1� are different sensitivity coeffi­
ðjÞ
as a monotonically increasing function and their formulations are cients; Δai ðtÞ represents acceleration difference between the
given below: lead vehicle j and subject vehicle i at the time t; αj , βj , and γj
� � are different weighting coefficients. This study used Equation
ðjÞ ðjþ1Þ ðjþN 1Þ ðjÞ
VOV ðΔXi ðtÞ; ΔXi ðtÞ; . . . ; ΔXi ðtÞÞ ¼ α1 V � ΔXi ðtÞ (15) for VOV ð:Þ. Moreover, they assumed αj as a decreasing
� � � � PN
ðjþ1Þ ðjþN 1Þ function with j, i.e. αj > αjþ1 and
þα2 V � ΔXi ðtÞ þ . . . þ αN V � ΔXi ðtÞ j¼1 αj ¼ 1. The functional
10 S. K. NIRMALE ET AL.

� � qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
form for αj is the same as that from Peng and Sun (2010). ð2�Þ ð2�Þ
Vsafe V2 ; ΔXi ¼ bTi þ b2 Ti2 þ V22 þ 2bΔXi (34)
Moreover, Li et al. (2011) illustrated that their model had
a greater stable region than that of the FVD model. Then, Vpred is used to calculate the safe speed for the subject vehicle
i as below:
Vi;safe ¼ bTi
Extensions of Intelligent Driver Model (IDM) rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi�ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi�ffiffi
� �
þ b2 Ti2 þ Vpre 2 þ 2b ΔX ð1Þ þ V
pred Ti min Vpred Ti ; ΔXconstant
To incorporate MVA, Treiber, Kesting, and Helbing (2006) con­ i

sidered cumulative stimuli between the subject vehicle i and lead (35)
vehicle j for N lead vehicles as given below:
where, a and b represent acceleration and deceleration, respectively.
0 12
2 The speed of the first and second lead vehicles are represented as V1
!β 3 ð1Þ
Vi ðtÞ XN B ΔXðj;desÞ ðtÞ C and V2 , respectively. ΔXi is the space gap between the first lead
41 5 B i C
ai ðtÞ ¼ amax
i amax
i BN C (29) ð2�Þ
ðdesÞ
Vi ðtÞ @ P ðjÞ A vehicle and subject vehicle i whereas ΔXi is the space gap between
j¼1 ΔX ðtÞ i the second lead vehicle and the first lead vehicle. The parameter ε
j¼1
represents fluctuations in units of acceleration (a).
P ðjÞ
where, Nj¼1 ΔXi ðtÞ represents net gaps between the subject vehi­
ðj;desÞ
cle i and lead vehicles j and ΔXi ðtÞ is represented below: Extensions of other models
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ðjÞ
ðj;desÞ Vi ðtÞ ðdesÞ Vi ðtÞΔVi ðtÞ Extensions of piece-wise linear car-following model
ΔXi ðtÞ ¼ C0 þ C1 ðdesÞ
þ V i ðtÞT i q ffi ffi ffi ffi ffi ffi ffi ffi ffi ffiffiffiffiffiffiffiffi A piece-wise linear car-following model was extended to incorpo­
Vi ðtÞ 2 amax bi
conf
i rate MVA in driver behavior (Farhi, Haj-Salem, and Lebacque
(30) 2012). It was a first-order discrete-time model, where speeds were
modeled as a function of spacings. In this study, the minimum form
Chen et al. (2010) showed that this formulation leads to an issue at was used rather than an additive form, which is commonly incor­
equilibrium flow; i.e. the space gap at equilibrium flow is dependent porated in the context of MVA-based driver behavior models.
on the number of lead vehicles leading to different desired head­ Moreover, the spacing to the jth lead vehicle was divided by j to
ways at the same uniform space gap at equilibrium. To overcome make sensitivities uniform across different lead vehicles. The math­
this drawback, Chen et al. (2010) proposed the following IDM- ematical formulation is presented below:
based model:
2 Xi ðt þ 1Þ ¼ Xi ðtÞ
!β 3 (
ðjÞ
! )
max 4 Vi ðtÞ 5 j 1 ΔXi ðtÞ
ai ðt þ Ti Þ ¼ ai 1 ðdesÞ
þ min ð1 þ λÞ min min αuw þ βuw
Vi ðtÞ 1�j�N u2U w2W j
!2
XN
ΔXi
ðjamÞ equ
ðtÞ þ Vi ðtÞTi (36)
max
ai αj (31)
j¼1
ΔXequ where, t þ 1 represents the next discrete time step, αuw and βuw , for
ðu; wÞ 2 U � W are parameters to be estimated, and U and W are
where, αj represents weight coefficients which decrease mono­ two finite sets of indices. With this formulation, this study con­
tonically as j increases to reflect that the influence of lead cluded that the minimum form used for considering more than one
vehicles decrease with an increase in the space gaps. They also lead vehicle and with a discounting factor ðλÞ used to favor the
concluded that the stable region increases after incorporat­ closest leader over distant ones. The proposed model was tested
ing MVA. only for single driver trajectory data. The authors reported para­
meter identification issues when a large number of vehicles (the
limit was not reported) were considered. Also, their modeling
Extension of collision-avoidance models framework did not account for heterogeneity in driving behavior.

Eissfeldt and Wagner (2003) extended a collision-avoidance-

based car-following model based on Gipps (1981), Krauss,
Wagner, and Gawron (1996) and Krauss (1998) and aimed at
clarifying the role of anticipation in the microscopic traffic model
through simulation and analytical calculations. The authors
assumed that a driver predicts the worst-case strategy (repre­
sented as Vpred ) that the immediate leader will choose in the
next time step. Vpred was a function of the leader’s desired
speed which in turn was a function of the speed of and spacing
with respect to the second lead vehicle. Considering that there is
a second lead vehicle in front of the first lead vehicle within
ð2Þ
a distance ΔX1 driving with speed V2 , then
Vpred ¼ maxfVdes
n � � o
ð2�Þ
Vdes ¼ min V1 þ a; V2;safe V2 ; ΔXi ; Vmax
Lane changing models
In lane changing, drivers judge the available gaps in the adjacent lane
and opt for the safest gap. Routinely, drivers interact with more than
one vehicle in the current and adjacent lanes by considering
dynamics (speed, spacing, etc.) of those vehicles in their decision-
making. Models describing the lane-changing behavior by design
incorporate multi-vehicle anticipation, and hence, justifiably, not
much special effort has been devoted in the literature to consider
MVA in lane changing models as is the case in single leader long­
itudinal movement models. Since some excellent review papers on
lane-changing models are already available (Moridpour, Sarvi, and
Rose 2010; Toledo 2007; Zheng 2014), we briefly review a few classi­
εa; 0g (32) cal studies on lane-changing models instead of a detailed review.
Lane-changing models can be broadly classified into the follow­
ing categories a) Gipps-type models, b) Utility theory-based mod­
(33)
els, c) Cellular automata-based models, d) Markov process-based
Table 3. Summary on models that incorporate multi-vehicle anticipation by drivers in heterogeneous disordered traffic streams.
Model Data utilized and
category Author(s) Proposed model region Strengths Weakness
h � � i h � �i
Extensions of Jin et al. ð1Þ ð2Þ ð1Þ ð2Þ –
full (2010)
ai ðtÞ ¼ α U ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ þ β G ΔVi ðtÞ; ΔVi ðtÞ • The proposed model takes into • The model assumed that only one
consideration the effect of the lateral side of the subject vehicle had
velocity gap between the subject vehicle and a laterally separated vehicle
difference the lead vehicle
(FVD) • The study demonstrates that
model incorporating the effect of the lateral
gap in the car-following model
stabilizes traffic flow, suppresses
h � � i h � �i traffic jams, and increases capacity
Li et al. ð1Þ ð2Þ ð3Þ ð1Þ ð2Þ ð3Þ –
(2015)
ai ðtÞ ¼ α U ΔXi ðtÞ; ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ þ β G ΔVi ðtÞ; ΔVi ðtÞ; ΔVi ðtÞ • The model considered a two-sided • Empirical evidence for the presence of
lateral gap as opposed to only one MVA behavior is not provided
side lateral gap • Models ignore behavioral differences
• Stability analysis of the proposed between different maneuvering
model revealed that the model is decisions (acceleration or
more efficient in dissipating deceleration)
perturbation than full velocity
difference-based models
Utility Choudhury Latent leader component • Trajectory data • Latent class approach is proposed to • Latent leader model is not an MVA-
Theory- and expðβj Zij ðtÞÞ
� j; j0 2 J ¼ FL; FD; FR collected at model acceleration decisions in weak based model, as it assumes that single
Pðli ðtÞÞ ¼ P j0 j0
Based Islam exp β Zi ðtÞ Mirpur Road of lane discipline traffic (unknown) leader
J
Models (2016) Dhaka, • The model does not model lateral
Acceleration
� component
� � �
ðjÞ ðjÞ ðjÞ Bangladesh movements
ai ðt þ Ti Þ ¼ Sen ΔXi ðtÞ Stim ΔVi ðtÞ þ εit
Nirmale, Discrete choice component • Vehicular • To incorporate multi-vehicle • The study does not consider the issue
Pinjari, Uij ¼ βTj Zij þ εij ; trajectory data anticipation, the concept of an of serial correlations that may arise
and collected at influence zone around a vehicle due to correlation across successive
Sharma j 2 ðAccelerate; Decelerate; MaintainconstantspeedÞ
Continuous component Chennai city, (subject vehicle) is introduced where observations of a vehicle.
(2021) India vehicles within the influence zone can • The study only considers rectangular-
mij ¼ αTj Kij þ ηij ; j 2 ðAccelerate; DecelerateÞ
potentially influence the subject shaped influence zone
vehicle’s driving behavior • The model only considers longitudinal
• Driving decisions were separated into movements
discrete and continuous
components – discrete component
for the decision to accelerate,
decelerate, or maintain a constant
speed and the continuous component
for the extent of acceleration or
deceleration
Nirmale, Discrete choice component • Vehicular • The study enhances the multi-vehicle • The study compares driving behavior
Pinjari, Uijt ¼ βTj Zijt þ ωij þ ψij þ εijt ; trajectory data anticipation-based discrete- in homogeneous and heterogeneous
and collected at continuous choice modeling disordered traffic streams using only
Sharma j 2 ðAccelerate; Decelerate; MaintainconstantspeedÞ
Continuous component Chennai city, framework by incorporating the one dataset from homogeneous
(2022) India influence of vehicle- and driver- traffic conditions and one dataset
mijt ¼ αTj Kijt � ωij þ �ij þ ηijt ; j 2 ðAccelerate; DecelerateÞ
• NGSIM data of specific unobserved factors on driver from heterogeneous disordered traffic
vehicle behavior streams
trajectories on • The study addresses the issue of serial • The study only considers rectangular-
a segment of correlations. shaped influence zone.
U.S. Highway • The study compares driving behavior • The model only considers longitudinal
TRANSPORTATION LETTERS

101, Los in homogeneous and heterogeneous movements

Angeles, United disordered traffic streams
States
(Continued)
11
 12
S. K. NIRMALE ET AL.

Table 3. (Continued).
Model Data utilized and
category Author(s) Proposed model region Strengths Weakness
Nirmale R
P
and Uijt ¼ βi0 þ βjr xir� þ �ij ; xir� : perceived traffic environment variable • Vehicular • The study addresses the issue of • Extent of acceleration is not modeled
r¼1 trajectory data parameter identification in MVA • The study only considers rectangular-
Pinjari collected at based driver behavior model shaped influence zone
(2023) Chennai city, • The empirical application of the • The model only considers longitudinal
India model provided interesting insights movements
into drivers’ perception errors. For
example, the perception errors are
greater for relative longitudinal
speeds than the longitudinal space
gaps and greater for vehicles
obliquely placed than those directly
ahead.
 TRANSPORTATION LETTERS 13

models, e) Hazard-based models, f) Fuzzy-logic-based models, and modeling driving behavior in HD traffic streams. Moreover, we
g) game theory-based models. Gipps (1986) was the first to model divide this section into two subsections − 1) studies those model
a driver’s lane change behavior. His model is based on the notion of only longitudinal movements and 2) studies those models’ two-
collision avoidance, treats lane changing as a deterministic process, dimensional movements.
and ignores inconsistency in driver behavior over time. Building on
Gipp’s model’s shortcomings, Yang and Koutsopoulos (1996)
developed and implemented a probabilistic lane-changing model Multi-vehicle anticipation-based driver behaviour model for
similar to Gipps’ model in the microscopic traffic simulator describing longitudinal movements
MITSIM. They defined lane changes as mandatory or discretionary
and modeled the lane changing process as four consecutive steps: Extensions of Full Velocity Difference (FVD) model
deciding on a lane change, selecting a target lane, examining an Jin et al. (2010) proposed a non-lane-based FVD model by taking
acceptable distance, and performing the lane shift. Another weak­ into consideration the effect of the lateral gap between the subject
ness of Gipps (1986) is the assumption that lane change occurs only vehicle and the lead vehicle. It was hypothesized that a driver is
when a large enough gap exists in the target lane. However, in heavy affected by the vehicle directly ahead as well as adjacent to it.
or congested traffic, this assumption would be unrealistic. To over­ h � � i
ð1Þ ð2Þ
come this limitation, Hidas (2002, 2005) proposed an improved ai ðtÞ ¼ α U ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ
modeling framework to capture the vehicular interaction induced h � �i
ð1Þ ð2Þ
by lane change, which was explicitly classified into three categories þ β G ΔVi ðtÞ; ΔVi ðtÞ (37)
based on observations from video-recording vehicular trajectory:
free, cooperative, and forced lane changes, to overcome this The following functional forms for U ð:Þ and Gð:Þ are employed in
limitation. their study:
Next, Ahmed et al. (1996) implemented utility theory to � � � �
model the decision process of lane change. The proposed ð1Þ ð2Þ
U ΔXi ðtÞ; ΔXi ðtÞ ¼ VOV ð1
ð1Þ ð2Þ
pi ÞΔXi ðtÞ þ pi ΔXi ðtÞ
model structure has four latent levels of decision hierarchy,
similar to steps given by Yang and Koutsopoulos (1996). (38)
Later, Toldeo (2003) provided an integrated modeling frame­
work where the car-following and lane-changing behavior mod­
� �
els were joined together in a single model. The integrated ð1Þ ð2Þ ð1Þ ð2Þ
G ΔVi ðtÞ; ΔVi ðtÞ ¼ ð1 pi ÞΔVi ðtÞ þ pi ΔVi ðtÞ (39)
model captured the trade-offs between the utility of being in
the correct lane and the speed advantage offered by a faster
lane. Furthermore, Kesting, Treiber, and Helbing (2007) pro­
posed a novel logic for modeling lane changing decisions based ð1Þ
LSi
on lane change’s anticipated advantages and disadvantages. For pi ¼ (40)
LSmax
example, the driver attempts to minimize overall braking
induced by lane changes. Recently, game theory has received
where, VOV ð:Þ is the optimal velocity function provided in Equation
plenty of attention to model lane change behavior. Talebpour, ð1Þ
Mahmassani, and Hamdar (2015) and Ali et al. (2019) are a few (15); pi captures the effect of lateral separation distance; LSi is the
notable attempts in this regard. lateral separation distance between the center line passing through
the subject vehicle i and the center line passing through the first
lead vehicle (i.e. j ¼ 1). LSmax is the maximum lateral separation
Models that incorporate multi-vehicle anticipation by distance beyond which the lead vehicle has no influence on the
drivers in heterogeneous disordered traffic streams subject vehicle and set to 3.6 m, i.e. typical lane width. Further,
Most driver behavior models assume that vehicles follow lane when LSi ¼ 0 (i.e. pi ¼ 0), the proposed model can be simplified as
discipline, follow the center line of the lane, and consider stimuli the FVD model. When LSi ¼ LSmax (i.e. pi ¼ 1), the first lead
solely from vehicles in front since they are developed for homo­ vehicle is on another lane, and the subject vehicle follows
geneous traffic conditions. However, in many developing coun­ the second lead vehicle. This study demonstrated that incorporating
tries (such as India, Bangladesh, China, etc.), lanes may not be the effect of the lateral gap in the car-following model stabilizes
well delineated, or lane discipline may not be effectively main­ traffic flow, suppresses traffic jams, and increases capacity.
tained, allowing vehicles to occupy any lateral position on the The above model assumed that only one side of the subject
road and encouraging two-dimensional movement for the vehi­ vehicle had a laterally separated vehicle. However, the subject
cles to navigate ahead. In addition, vehicles from the side can cut- vehicle may be influenced by vehicles traveling laterally to either
in to the front of a subject vehicle anytime. Furthermore, HD side of it. Considering this, Li et al. (2015) developed a two-sided
traffic streams comprise a wide variety of vehicle classes (such as lateral gap FVD model for non-lane discipline traffic as given
passenger cars, motorbikes, buses, trucks, three-wheeled auto- below:
rickshaws, and non-motorized vehicles) with considerably differ­ h � � i
ð1Þ ð2Þ ð3Þ
ent physical and operational characteristics. Most of these classes ai ðtÞ ¼ α U ΔXi ðtÞ; ΔXi ðtÞ; ΔXi ðtÞ Vi ðtÞ
have substantial representation in the traffic streams. In contrast, h � �i
ð1Þ ð2Þ ð3Þ
homogeneous traffic streams are dominated mostly by passenger þ β G ΔVi ðtÞ; ΔVi ðtÞ; ΔVi ðtÞ (41)
cars having similar physical and operational characteristics. Such
distinctive characteristics of HD traffic streams cause differences It was assumed that the first ðj ¼ 1Þ, second ðj ¼ 2Þ, and third lead
in the driver behavior between homogeneous and HD traffic ðj ¼ 3Þ vehicles are traveling on the right front side, left front side,
streams (Nirmale, Pinjari, and Sharma 2022). The same can be and immediate front of the subject vehicle, respectively.
reflected in MVA behavior. Therefore, in this section, we mainly Accordingly, stimuli were calculated for each of the lead vehicles.
focus on the studies that have considered the MVA effect while The following functional forms for Uð:Þ and Gð:Þ were utilized:
14 S. K. NIRMALE ET AL.
� �
ð1Þ ð2Þ ð3Þ
U ΔXi ðtÞ; ΔXi ðtÞ; ΔXi ðtÞ
8 h i
< ð1Þ ð3Þ ð1Þ
V ð1 2pi ÞΔXi ðtÞ þ 2pi ΔXi ðtÞ "LSi 2 ð0; 0:5LSmax Þ
¼ h i multi-vehicle anticipation-based discrete-continuous choice mod­
: V ð2pi 1ÞΔX ð2Þ ðtÞ þ 2ð1 pi ÞΔX ð3Þ ðtÞ "LSð1Þ 2 ð0:5LSmax ; LSmax Þ
i i i eling framework. Specifically, the modeling framework was
� �
ð1Þ ð2Þ ð3Þ
G ΔVi ðtÞ; ΔVi ðtÞ; ΔVi ðtÞ
( correlation issue by employing a simple empirical strategy. Using
ð1Þ ð3Þ ð1Þ
ð1 2pi ÞΔXi ðtÞ þ 2pi ΔXi ðtÞ"LSi 2 ð0; 0:5LSmax Þ
¼ ð2Þ ð3Þ ð1Þ
ð2pi 1ÞΔXi ðtÞ þ 2ð1 pi ÞΔXi ðtÞ"LSi 2 ð0:5LSmax ; LSmax Þ
LS1i
pi ¼
LSmax
Similar to Jin et al. (2010), this study also employed the optimal
velocity function as per Equation (15). Stability analysis of the
proposed model revealed that the model is more efficient in dis­
sipating perturbation than previous FVD-based models.
Utility theory-based models
In HD traffic streams, the subject vehicle is surrounded by many
vehicles, resulting in multiple stimuli sources from multiple vehi­
cles. Therefore, drivers might react to stimuli from a governing
leader. However, only the driver’s final actions (such as applied
acceleration) are observed from trajectory data, and the governing
leader is latent to the analyst. This prompts the need to develop
a modeling framework to consider the latent leader while analyzing
driver behavior. To do so, a latent leader approach was proposed by
Choudhury and Islam (2016) to model acceleration decisions.
Particularly, a random utility-based modeling framework was pro­
posed with two components: latent leader component and accel­
eration component. The former was modeled as a random utility-
based discrete choice model, and the probability of the front lead
vehicle l (which can be front left, front direct, front right) being
a governing leader of the subject vehicle i was expressed as:
� �
j
exp βj Zi ðtÞ
Pðli ðtÞÞ ¼ P � 0 0 � j; j0 2 J ¼ FL; FD; FR (45)
j
exp βj Zi ðtÞ
J
j
where, Zi ðtÞrepresents
the explanatory variables associated with the
j
lead vehicle j and β is an estimated parameter vector associated
with respect to the lead vehicle j. Whereas the acceleration compo­
nent was modeled using the GM model. Strictly speaking, the latent
leader model is not an MVA-based model, as it assumes that single
(unknown) leader.
More recently, Nirmale, Pinjari, and Sharma (2021) proposed
an MVA-based discrete-continuous choice modeling framework
for describing driver behavior in HD traffic conditions. To incor­
porate MVA, they introduced the concept of an influence zone
around a vehicle (subject vehicle) where vehicles within the
influence zone can potentially influence the subject vehicle’s
driving behavior. Moreover, driving decisions were separated
into discrete and continuous components – discrete component
for the decision to accelerate, decelerate, or maintain a constant
speed and the continuous component for the extent of accelera­
tion or deceleration. The estimation results from this study
underscored the importance of considering the MVA effect for
describing driving behavior in HD traffic conditions. Specifically,
their study concluded that drivers in HD traffic conditions not
only consider vehicles that are ahead of their vehicle but also
consider those vehicles that are on either side. Also, they found
that the extent of influence of stimuli with respect to lead vehicles
is different on discrete decisions and continuous decisions.
Subsequently, Nirmale, Pinjari, and Sharma (2022) proposed
a panel data version of Nirmale, Pinjari, and Sharma (2021)’s
(42) enhanced by incorporating the influence of vehicle- and driver-
specific unobserved factors on driver behavior. In addition,
Nirmale, Pinjari, and Sharma (2022) study addressed the serial
this proposed model, driving behavior in HD and homogeneous
(43)
traffic streams were compared by estimating empirical models
using trajectory data from Chennai, India (HD traffic) and recon­
structed NGSIM data (homogeneous traffic).
(44) In another study, Nirmale and Pinjari (2023) incorporated dri­
ver errors in perceiving stimuli from vehicles directly ahead, obli­
quely placed, and on either side in their driver behavior model. The
empirical application of their model provided interesting insights
on drivers’ perception errors. For example, the perception errors
are greater for relative longitudinal speeds than the longitudinal
space gaps, and greater for vehicles obliquely placed than those
directly ahead.
Driver behaviour model for describing two-dimensional
movement
Models that can describe simultaneous lateral and longitudinal
movements offer more realism in describing HD traffic. Yet, such
studies are rare (Mahapatra, Maurya, and Chakroborty (2018) and
Chakroborty, Maurya, and Vikram (2019) provide a review of such
efforts). In the ensuing paragraphs, a brief discussion of these
models is provided, along with some recent studies that are not
included in previous reviews.
Chakroborty, Agrawal, and Vasishtha (2004) developed
a comprehensive microscopic model for two-way traffic using
a potential field approach. The proposed model had two major
components, namely, (1) Steering Response Model (SRM, for pre­
dicting the steering angles adoption with time) and (2) Acceleration
Response Model (ARM, to predict the rate of acceleration and
deceleration over time). Later, a Comprehensive Unidirectional
Traffic Simulation Model (CUTSiM) was proposed by Maurya
(2007) that considered the Indian traffic conditions explicitly. The
proposed model included a lateral control model component that
describes the driver’s decision to choose a suitable steering angle
based on the hypothesized best path along with its longitudinal
control model. However, this model was not extensively calibrated
and validated. Furthermore, Kanagaraj and Treiber (2018) pro­
posed a two-dimensional time-continuous model for the mixed
traffic flow of motorized and non-motorized vehicles based on the
force-field model. The trajectory data from Chennai, India, was
utilized to calibrate this model. Recently, Delpiano et al. (2019)
developed a two-dimensional microscopic car-following model
using the social force approach. They argued that the distance
maintained by the driver to avoid collisions in all directions is
a critical factor and thus, proposed the multi-directional collision
avoidance behavior model where two-dimensional repulsive force
between vehicles was modeled. Specifically, three forces were con­
sidered that act on each vehicle: the acceleration force (willingness
to accelerate), lane force (tendency to be in the center in a specific
lane), and the repulsive force for collision avoidance. Simulation
experiments were also performed to reproduce two-dimensional
collision avoidance behavior. They concluded that the proposed
model is a sound starting point for building autonomous vehicles
traffic flow models and can improve autonomous driving
algorithms.

Next, Mathew, Munigety, and Bajpai (2013) proposed a strip-
based approach for the simulation of HD traffic conditions. They
developed a simulator named Simulation of Mixed Traffic Mobility
using a traditional lane-based simulator SUMO. In the proposed
approach, the road was divided into thin strips allowing continuous
lateral movement rather than conventional discrete lane changing.
They observed that the reduction of strip width increases the
throughput indicating improved utilization of road space.
Furthermore, Lee, Polak, and Bell (2009) proposed an agent-
based model to simulate motorcycle behavior in HD traffic condi­
tions. They developed three models to mimic motorcycle move­
ment patterns, namely, the longitudinal headway model, the
oblique and lateral headway model, and the path choice model.
The longitudinal headway model described the motorcyclist driving
behavior that they maintain a shorter headway when aligning to the
edge of the lead vehicle. The oblique and lateral headway model
described the headway distribution of motorcycles when they are
following the lead vehicles obliquely.
The virtual lane-based movement of motorcycles was modeled as
a multinomial logit. Next, Shiomi et al. (2012) proposed a utility
theory-based approach to describe the two-dimensional movement
of the two-wheelers. This approach captured the characteristics of
driver perception of the surrounding traffic situation albeit, it failed
to capture the heterogeneity across the drivers. Following a similar
utility theory-based approach, Sarkar et al. (2020) proposed a modeling
framework that models driver’s two-dimensional movement behavior.
This framework proposed two components: area selection and vehicle
movement. In the area selection component, the possible movements
of subject vehicles for next time steps were considered, and two-
dimensional space ahead of the subject vehicle was divided into the
number of realistic cones and treated as choice alternatives for the
subject vehicle. A multinomial logit model was used to model the
driver’s discrete decisions. For the vehicle movement component,
a modified intelligent driver model was proposed, which simulates
the subject vehicle’s next position along the selected direction. Both
components of the proposed framework were calibrated separately
using a real trajectory data set collected at Chennai, India.
Recently, Amrutsamanvar (2020) analyzed the lateral movement
decisions of two-wheeler drivers in HD traffic conditions using
different machine-learning (ML) models. The study framed the
lateral movement decision of two-wheeler drivers as a multiclass
classification problem, where the driver must select one alternative
from a set of three discrete options: turn left, turn right, or move
straight. The study used vehicle trajectory data from Chennai,
India, to train the ML models. The study revealed that the lateral
distance to a ready-to-overtake position is more important than the
longitudinal gap between the lead and subject vehicles in influen­
cing the lateral movement decisions of two-wheeler drivers.
Critical findings
This paper provides a critical review of notable attempts to
incorporate MVA in driving behavior models. Primarily, MVA-
based driver behavior models follow a similar pattern of extend­
ing the single-leader car-following model by adding weighted
stimuli from vehicles ahead. Interestingly, the influence of only
a fixed number (say 3 or 5) of vehicles ahead is considered in
these models. Moreover, vehicles far ahead have a low influence
on drivers’ maneuvering decisions, and hence, weights on the
stimuli decrease as the distance from the subject vehicle increases.
Furthermore, the type of vehicle ahead (a car or a truck) also
influences the MVA behavior of drivers and, thereby, their man­
euvering decisions. For example, a car driver following a truck is
blind to vehicles ahead of the truck (second leader, third leader,
TRANSPORTATION LETTERS 15
etc.) and, thus, gives low to no weightage to stimuli from those
vehicles. It has also been demonstrated that drivers not only react
to the vehicles directly ahead but also those that are placed
obliquely ahead or on either side of the subject vehicle, specifi­
cally in HD traffic streams. Furthermore, incorporating percep­
tion errors in MVA-based driver behavior models improved the
behavioral soundness of these models. Nirmale and Pinjari (2023)
found that the stimuli perception errors are greater for vehicles
that are obliquely placed than vehicles that are directly ahead.
This may be because drivers pay greater attention to vehicles
directly ahead than those obliquely placed. In addition, all studies
have concluded that MVA-based models perform better in stabi­
lizing the traffic and suppressing traffic oscillations than single-
lead vehicle-based models. Treiber, Kesting, and Helbing (2006)
revealed that the model with MVA has a higher threshold for
traffic stability over the single leader model, i.e. traffic remains
stable for higher values of reaction times for the model with
MVA, which is consistent with real-world traffic observations.
Most of the studies were restricted to three lead vehicles ahead of
the subject vehicle and that too those that were directly ahead of the
subject vehicle. The number of lead vehicles to be considered in the
MVA framework may depend on the type of lead vehicle as con­
jectured by Hoogendoorn and Ossen (2006). In addition, surround­
ing vehicles that are not only straight ahead of the driver but also
obliquely placed influence drivers’ decisions (Nirmale, Pinjari, and
Sharma 2021). Specifically, in the HD traffic streams, it becomes
more important to consider vehicles around the subject vehicle
because of its typical characteristics (see Section 4). Surprisingly, the
influence of vehicle type is largely ignored by all studies reviewed here
except two (Hoogendoorn and Ossen 2006; Nirmale, Pinjari, and
Sharma 2021). In fact, the number of vehicles considered by the
driver is always kept constant irrespective of the type of vehicle ahead.
Modeling requirements, challenges and future research
directions
In this section, we discuss common issues in the present MVA-based
driver behavior modeling framework and future research needs.
How many vehicles ahead to consider in MVA-based driver
behavior models?
As discussed earlier, the influence of only a fixed number (say 3
or 5) of vehicles ahead is considered in these models. Hence, future
endeavors can focus on investigating the impact of considering the
different number of vehicles ahead in both homogeneous and HD
traffic streams on the numerical and behavioral soundness of the
model. Similarly, the influence of side vehicle’s impact on driving
behavior can be examined leading to modifications in the existing
MVA-based models.
In the recent study by Nirmale, Pinjari, and Sharma (2021),
a concept of an influence zone was proposed, and it was assumed
that vehicles within the influence zone (irrespective of the number
and position with respect to the subject vehicle) might impact the
microscopic driving behavior of the driver. One can employ this
approach to incorporate MVA. One of the limitations of the
Nirmale, Pinjari, and Sharma (2021) study is that they used fixed
size and shape influence zone for all drivers. However, the shape
and size might be different for drivers and different vehicle types
of the subject vehicle. Hence, future research endeavors can
explore different shapes and sizes for different drivers and vehicle
types.
16 S. K. NIRMALE ET AL.
Additional critical factors to be considered in MVA-based
driver behavior models
Influence of vehicle type
The traffic stream consists of different categories of vehicles, includ­
ing cars, motorbikes, buses, heavy vehicles, etc. The relative per­
centages of these vehicle types change depending on the type of
stream (homogeneous or HD), type of facility (urban streets or
highways), geographical locations (hilly terrain vs plain terrain),
etc. MVA also depends on the type of vehicle. Let’s first consider the
type of subject vehicle. The distance up to which vehicles ahead will
influence the driver will be larger for bus/truck drivers than car
drivers because the eye level of bus/truck drivers is at a greater
height; thus, can see for a larger distance as compared to car drivers
leading to differences in their MVA behavior. Second, the type of
vehicles in the visual field also influences the MVA behavior.
Imagine a situation that a car is following a bus/truck. Vehicles
ahead of the bus/truck will not be visible to the car driver; thus,
stimulus from those vehicles shall not be considered in the car
driver’s decision-making. Surprisingly, this factor is largely ignored
by all studies reviewed here except one (Hoogendoorn and Ossen
2006). In fact, the number of vehicles considered by the driver is
always kept constant irrespective of the type of vehicle ahead. For
realistic modeling of MVA, the type of vehicle both at the subject
vehicle level and at the traffic stream ahead shall be considered.
Drivers’ useful visual field
As per Mackworth (1965), ‘the useful visual field can be defined
as the area around the fixation point (the point in space on
which the eyes are focused) inside which information can be
perceived.’ If the quantity of information to be processed
increases, the useful visual field decreases (Rogé et al. 2004).
For example, the useful visual field decreases if the number of
vehicles within the visual field increases. Moreover, it
decreases with the increase in speed. This indicates that the
useful visual field will have a strong bearing on the MVA
behavior of drivers because it directly influences the stimulus
perception by drivers. Therefore, future endeavors on building
MVA-based driver behavior models shall consider this impor­
tant factor.
Human factors other than multi-vehicle anticipation
Human factors such as socioeconomic characteristics (e.g. age,
gender, etc.), driver’s reaction time, estimation and perception
error, driving style, driver’s risk-taking behavior, distraction etc.
(Sharma et al. 2017; Treiber and Kesting 2013), are generally
ignored when mimicking driver behavior. These human factors
contribute to driver errors which is the number one cause of on-
road accidents (WHO 2018). Not surprisingly, apart from reaction
time, other human factors are not included in models discussed in
this study, thus, making them less suitable to describe driver beha­
vior in crash-prone and risk-taking situations. Therefore, for
a richer representation of the cognitive process behind decision-
making during such events and others it is imperative to incorpo­
rate human factors (enumerated above) in the current modeling
framework (Saifuzzaman and Zheng 2014).
It is worth noting that incorporating human factors may further
complicate the model in terms of the number of parameters that one
has to consider. Also, this might lead to parameter identification issues.
Hence, careful thinking and formal analysis are needed before specify­
ing models with different human factors. In the end, there is a trade-off
between models’ explanatory power and complexity in terms of the
number of parameters to be estimated/calibrated.
Another important point is that the above enumerated human
factors remain mainly unobserved because the data collection pro­
cess for building, calibrating, and validating driver behavior models
majorly focuses on gathering the trajectory data from videos or GPS
devices. Another reason could be that some of these factors are
unknown to the analyst. Little effort has been made to acknowledge
drivers’ unobserved factors and their implications on the modeling
framework. A recent study by Nirmale, Pinjari, and Sharma (2022)
incorporated drivers’ unobserved factors latent to analysts to better
mimic driver behavior in HD traffic streams. In future endeavors,
considering the effect of human factors explicitly or as unobserved
factors is important to improve driver behavior models.
Incorporating the aforementioned factors will make MVA-based
driver behavior models more realistic.
Data needs for calibration and validation of MVA based driver
behavior models
We did not find studies focussing on collecting data specifically to
capture multi-vehicle anticipation behavior, probably because most
of the time MVA-based models are simple extensions of single-lead
vehicle-based models. As revealed in this study, MVA is not
a straightforward process. Before incorporating it, one needs to
answer how many vehicles a driver considers? What are the driver’s
focus areas? How to quantify a driver’s useful visual field? How does
the useful visual field vary for a particular driver? How does MVA
change based on the type of subject vehicle (bus or car or motor­
bike) and on vehicle classes in front? To answer all these questions,
it is important to collect detailed driver behavior data. For instance,
researchers shall focus on collecting data up to a trap length of 100
m to 500 m so that kinematics of a few vehicles ahead can be
captured. Notably, driving simulators offer various advantages,
specifically the collection of detailed driver-level data. Various
scenarios with a varying number of vehicles ahead, different vehicle
classes, varied placement of vehicles in the driving scene, etc., can
be created and tested. Next, eye trackers can be utilized to identify
driver focus points within the visual field. Combining driving
simulators and eye-tracking data can assist in a better understand­
ing and modeling of the MVA behavior.
The issue of parameter identification in MVA based driver
behavior models
MVA-based driver behavior models include more parameters when
compared to the single leader driver behavior models, one needs to
investigate whether the new parameters are identifiable and thereby
contribute to the model output. Such analysis will assist in deciding
what parameters need to be fixed or calibrated or removed from the
model. There are two approaches to examine model identification –
theoretical and empirical. In theoretical identification, specific rules
developed for generalized models are applied to see whether the
model is identifiable. For example, rank and order conditions are
assessed to determine whether the specified choice model is identi­
fiable (Walker 2001) as employed by Nirmale and Pinjari (2023).
On the other hand, in empirical identification, model outputs or
model-based performance statistics (such as likelihood) are exam­
ined using numerical simulation experiments. For example, asses­
sing that the parameters converge to the same point and likelihood
given different starting values, and verifying that the Hessian of the
log-likelihood function is non-singular via Monte Carlo experi­
ments (Walker 2001), also performed by Nirmale and Pinjari
(2023). Another method is sensitivity analysis in which the aim is

to inspect the contribution of each parameter to model output (Lu,
Ni, and Washburn 2015; Punzo, Montanino, and Ciuffo 2014).
After recognizing identifiable parameters, analysts can fix the uni­
dentifiable parameters to a particular value or may decide to remove
them and then estimate/calibrate their proposed model. In a nutshell,
different methods are available to analyze parameter identification.
Still, the application of those methods is sparse in the existing literature.
Derivation of macroscopic properties of MVA-based driver
behavior models
The capability to reproduce the macroscopic traffic flow pattern, such
as realistic relationships among flow, speed, and density, is a critical
property to assess when analyzing driver behavior models. However,
limited efforts are made to investigate the macroscopic traffic flow
properties resulting from MVA-based driver behavior models. It may
not be easy to do so using analytic approaches. Therefore, a limited
number of studies (Eissfeldt and Wagner 2003; Lenz, Wagner, and
Sollacher 1999; Nirmale 2022) aggregated the simulated vehicle trajec­
tories from MVA-based driver behavior models to derive the resulting
macroscopic traffic flow properties. The analytical derivation of the
macroscopic properties of MVA-based driver behavior models is an
important direction for future research.
Conclusions
This paper provides a review and synthesis of MVA-based
driver behavior models in the literature for homogeneous
and heterogeneous disordered (HD) traffic streams. None of
the previous review articles on driver behavior models exam­
ined MVA-based driver behavior models in detail. Therefore,
this paper systematically organizes the different approaches
used by researchers to capture MVA behavior in both homo­
geneous and HD traffic streams. All existing models have been
evaluated critically and summarized in Tables 2 and 3.
Overall, our findings indicate that MVA-based driver beha­
vior models available in the literature follow a similar pattern
of extending the established single-leader car-following mod­
els, considering vehicles that are directly ahead (in the same
lane) and focussing on a fixed number of vehicles ahead. For
HD traffic streams, drivers also consider stimuli from vehicles
obliquely placed or on either side. Furthermore, the type of
vehicle ahead (a car or a truck) also influences the MVA
behavior of drivers and, thereby, their maneuvering decisions.
Furthermore, it is observed that incorporating perception
errors in MVA-based driver behavior models improved the
behavioral soundness of these models. Additionally, it is
found that the stimuli perception errors are greater for vehi­
cles that are obliquely placed than vehicles that are directly
ahead. In addition, all studies have concluded that MVA-based
models perform better in stabilizing the traffic and suppressing
traffic oscillations than single-lead vehicle-based models.
Next, this review discusses modeling requirements, chal­
lenges and future research directions, such as how many vehi­
cles ahead shall be considered when incorporating the MVA
behavior, what additional factors (influence of vehicle type,
driver’s visual field, human factors other than multi-vehicle
anticipation) shall be included to make models more realistic,
data requirements for calibration and validation, the issue of
parameter identification, and the derivation of macroscopic
properties of these models. To this end, this review is
a useful step toward gathering and organizing what we already
know, figuring out where more research is needed, and making
suggestions for analyzing driver behavior in the future.
TRANSPORTATION LETTERS 17
Note
1. Different studies use different functional forms for the optimal velocity
function. To avoid confusion and notation burden, we use a generic
symbol VOV ð:Þ for the optimal velocity function throughout the paper.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Funding
The work was supported by the Ministry of Education (IN) .
ORCID
Sangram Nirmale http://orcid.org/0000-0001-7930-8195
References
Ahmed, K., M. Ben-Akiva, H. Koutsopoulos, and R. Mishalani. 1996. “Models of
Freeway Lane Changing and Gap Acceptance Behavior.” Transportation and
Traffic Theory 13:501–515.
Ali, Y., Z. Zheng, M. M. Haque, and M. Wang. 2019. “A Game Theory-Based
Approach for Modelling Mandatory Lane-Changing Behaviour in
a Connected Environment.” Transportation Research Part C: Emerging
Technologies 106:220–242. https://doi.org/10.1016/j.trc.2019.07.011 .
Amrutsamanvar, R. 2020. “Modeling Lateral Movement Decisions of Powered
Two Wheelers in Disordered Heterogeneous Traffic Conditions.”
Transportation Letters 14 (3): 1–20. https://doi.org/10.1080/19427867.2020.
1839718.
Arun, A., M. M. Haque, A. Bhaskar, S. Washington, and T. Sayed. 2021.
“A Systematic Mapping Review of Surrogate Safety Assessment Using
Traffic Conflict Techniques.” Accident Analysis & Prevention 153:106016.
https://doi.org/10.1016/j.aap.2021.106016.
Asaithambi, G., V. Kanagaraj, and T. Toledo. 2016. “Driving Behaviors: Models
and Challenges for Non-Lane Based Mixed Traffic.” Transportation in
Developing Economies 2 (2). https://doi.org/10.1007/s40890-016-0025-6.
Azam, M. S., A. Bhaskar, and M. M. Haque. 2020. “Driving Behaviour Modelling
in the Context of Heterogeneous Traffic and Poor Lane Discipline
Conditions: The State of the Art and Beyond.” Transportmetrica A:
Transport Science 18 (3): 1–68.
Bando, M., K. Hasebe, K. Nakanishi, and A. Nakayama. 1998. “Analysis of
Optimal Velocity Model with Explicit Delay.” Physical Review E 58 (5):
5429. https://doi.org/10.1103/PhysRevE.58.5429.
Bando, M., K. Hasebe, A. Nakayama, A. Shibata, and Y. Sugiyama. 1995.
“Dynamical Model of Traffic Congestion and Numerical Simulation.”
Physical Review E 51 (2): 1035–1042. https://doi.org/10.1103/PhysRevE.51.
1035.
Bexelius, S. 1968. “An Extended Model for Car-Following.” Transportation
Research 2 (1): 13–21. https://doi.org/10.1016/0041-1647(68)90004-X.
Brackstone, M., and M. McDonald. 1999. “Car-Following: A Historical Review.”
Transportation Research Part F: Traffic Psychology and Behaviour 2 (4):
181–196. https://doi.org/10.1016/S1369-84780000005-X .
Chakroborty, P., S. Agrawal, and K. Vasishtha. 2004. “Microscopic Modeling of
Driver Behavior in Uninterrupted Traffic Flow.” Journal of Transportation
Engineering 130 (4): 438–451. https://doi.org/10.1061/(ASCE)0733-947X
(2004)130:4(438).
Chakroborty, P., A. K. Maurya, and D. Vikram. 2019. “Understanding and
Modelling Disorderly Traffic Streams.” Journal of the Indian Institute of
Science 99 (4): 1–13. https://doi.org/10.1007/s41745-019-00135-7.
Chandler, R. E., R. Herman, and E. W. Montroll. 1958. “Traffic Dynamics:
Studies in Car Following.” Operations Research 6 (2): 165–184. https://doi.
org/10.1287/opre.6.2.165.
Chen, J., R. Liu, D. Ngoduy, and Z. Shi. 2016. “A New Multi-Anticipative
Car-Following Model with Consideration of the Desired Following
Distance.” Nonlinear Dynamics 85 (4): 2705–2717. https://doi.org/10.1007/
s11071-016-2856-4.
Chen, J., Z. Shi, and Y. Hu. 2012. “Stabilisation Analysis of a Multiple
Look-Ahead Model with Driver Reaction Delays.” International Journal of
Modern Physics C 23 (6): 1250048. https://doi.org/10.1142/
S0129183112500489.
Chen, X.-Q., W.-J. Xie, J. Shi, and Q.-X. Shi. 2010. “Perturbation and Stability
Analysis of the Multi-Anticipative Intelligent Driver Model.” International
18 S. K. NIRMALE ET AL.
Journal of Modern Physics C 21 (5): 647–668. https://doi.org/10.1142/
S0129183110015397.
Choudhury, C. F., and M. M. Islam. 2016. “Modelling Acceleration Decisions in
Traffic Streams with Weak Lane Discipline: A Latent Leader Approach.”
Transportation Research Part C: Emerging Technologies 67:214–226. https://
doi.org/10.1016/j.trc.2016.02.010.
Das, S., and A. K. Maurya. 2018. “Modelling of Motorised Two-Wheelers:
A Review of the Literature.” Transport Reviews 38 (2): 209–231. https://doi.
org/10.1080/01441647.2017.1317049.
Delpiano, R., J. Herrera, J. Laval, and J. Coeymans-Avaria. 2019. “A
Two-Dimensional Car-Following Model for Two-Dimensional Traffic Flow
Problems.” Transportation Research Part C: Emerging Technologies
114:504–516. https://doi.org/10.1016/j.trc.2020.02.025.
Eissfeldt, N., and P. Wagner. 2003. “Effects of Anticipatory Driving in a Traffic Flow
Model.” The European Physical Journal B-Condensed Matter and Complex
Systems 33 (1): 121–129. https://doi.org/10.1140/epjb/e2003-00149-8 .
Farhi, N., H. Haj-Salem, and J. P. Lebacque. 2012. “Multianticipative
Piecewise-Linear Car-Following Model.” Transportation Research Record:
Journal of the Transportation Research Board 2315 (1): 100–109. https://doi.
org/10.3141/2315-11.
Gazis, D. C., R. Herman, and R. B. Potts. 1959. “Car-Following Theory of
Steady-State Traffic Flow.” Operations Research 7 (4): 499–505. https://doi.
org/10.1287/opre.7.4.499.
Gazis, D. C., R. Herman, and R. W. Rothery. 1961. “Nonlinear Follow-The-
Leader Models of Traffic Flow.” Operations Research 9 (4): 545–567. https://
doi.org/10.1287/opre.9.4.545.
Ge, H. X., S. Q. Dai, L. Y. Dong, and Y. Xue. 2004. “Stabilisation Effect of Traffic
Flow in an Extended Car-Following Model Based on an Intelligent
Transportation System Application.” Physical Review E 70 (6): 066134.
https://doi.org/10.1103/PhysRevE.70.066134.
Gipps, P. G. 1981. “A Behavioural Car-Following Model for Computer
Simulation.” Transportation Research Part B: Methodological 15 (2):
105–111. https://doi.org/10.1016/0191-2615(81)90037-0.
Gipps, P. G. 1986. “A Model for the Structure of Lane-Changing Decisions.”
Transportation Research Part B: Methodological 20 (5): 403–414. https://doi.
org/10.1016/0191-2615(86)90012-3.
Hasebe, K., A. Nakayama, and Y. Sugiyama. 2003. “Dynamical Model of
a Cooperative Driving System for Freeway Traffic.” Physical Review E
68 (2): 026102. https://doi.org/10.1103/PhysRevE.68.026102.
Helly, W. 1959. “Simulation of Bottlenecks in Single-Lane Traffic Flow.”
Proceedings of the Symposium on Traffic Flow Theory, Research
Laboratories, General Motors 207–238.
Hidas, P. 2002. “Modelling Lane Changing and Merging in Microscopic Traffic
Simulation.” Transportation Research Part C: Emerging Technologies 10 (5–
6): 351–371. https://doi.org/10.1016/S0968-090X(02)00026-8.
Hidas, P. 2005. “Modelling Vehicle Interactions in Microscopic Simulation of
Merging and Weaving.” Transportation Research Part C: Emerging
Technologies 13 (1): 37–62. https://doi.org/10.1016/j.trc.2004.12.003.
Hoogendoorn, S., and S. Ossen. 2006. “Empirical Analysis of Two-Leader
Car-Following Behavior.” European Journal of Transport and Infrastructure
Research 6 (3): 229–246.
Hoogendoorn, S., S. Ossen, and M. Schreuder. 2006. “Empirics of
Multianticipative Car-Following Behavior.” Transportation Research Record
1965 (1): 112–120. https://doi.org/10.1177/0361198106196500112.
Hu, Y., T. Ma, and J. Chen. 2014. “An Extended Multi-Anticipative Delay Model
of Traffic Flow.” Communications in Nonlinear Science and Numerical
Simulation 19 (9): 3128–3135. https://doi.org/10.1016/j.cnsns.2014.02.006.
Jiang, R., Q. Wu, and Z. Zhu. 2001. “Full Velocity Difference Model for a
Car-Following Theory.” Physical Review E 64 (1): 017101. https://doi.org/
10.1103/PhysRevE.64.017101.
Jin, S., D. Wang, P. Tao, and P. Li. 2010. “Non-Lane-Based Full Velocity Difference
Car Following Model.” Physica A: Statistical Mechanics and Its Applications
389 (21): 4654–4662. https://doi.org/10.1016/j.physa.2010.06.014.
Jin, Y., M. Xu, and Z. Gao. 2011. “KdV and Kink-Antikink Solitons in an
Extended Car-Following Model.” Journal of Computational and Nonlinear
Dynamics 6 (1): 011018. https://doi.org/10.1115/1.4002336.
Kanagaraj, V., and M. Treiber. 2018. “Self-Driven Particle Model for Mixed
Traffic and Other Disordered Flows.” Physica A: Statistical Mechanics and Its
Applications 509:1–11. https://doi.org/10.1016/j.physa.2018.05.086.
Kesting, A., M. Treiber, and D. Helbing. 2007. “General Lane-Changing Model
MOBIL for Car-Following Models.” Transportation Research Record: Journal
of the Transportation Research Board 1999 (1): 86–94. https://doi.org/10.
3141/1999-10.
Krauss, S. 1998. Microscopic modeling of traffic flow: Investigation of collision
free vehicle dynamics. PhD thesis.
Krauss, S., P. Wagner, and C. Gawron. 1996. “Continuous Limit of the
Nagel-Schreckenberg Model.” Physical Review E 54 (4): 3707. https://doi.
org/10.1103/PhysRevE.54.3707.
Lee, T.-C., J. W. Polak, and M. G. Bell. 2009. “New Approach to Modeling Mixed
Traffic Containing Motorcycles in Urban Areas.” Transportation Research
Record 2140 (1): 195–205. https://doi.org/10.3141/2140-22.
Lenz, H., C. Wagner, and R. Sollacher. 1999. “Multi-Anticipative Car-Following
Model.” The European Physical Journal B-Condensed Matter and Complex
Systems 7 (2): 331–335. https://doi.org/10.1007/s100510050618 .
Li, Z. P., and Y. C. Liu. 2006. “Analysis of Stability and Density Waves of Traffic
Flow Model in an ITS Environment.” The European Physical Journal
B-Condensed Matter and Complex Systems 53 (3): 367–374. https://doi.org/
10.1140/epjb/e2006-00382-7.
Li, Y., D. Sun, W. Liu, M. Zhang, M. Zhao, X. Liao, and L. Tang. 2011. “Modeling
and Simulation for Microscopic Traffic Flow Based on Multiple Headway,
Velocity and Acceleration Difference.” Nonlinear Dynamics 66 (1–2): 15–28.
https://doi.org/10.1007/s11071-010-9907-z.
Li, Y., L. Zhang, S. Peeta, H. Pan, T. Zheng, Y. Li, and X. He. 2015. “Non-Lane-
Discipline-Based Car-Following Model Considering the Effects of Two-Sided
Lateral Gaps.” Nonlinear Dynamics 80 (1): 227–238. https://doi.org/10.1007/
s11071-014-1863-6.
Lu, B., S. Ni, and S. S. Washburn. 2015. “A Support Vector Regression Approach
for Investigating Multianticipative Driving Behavior.” Mathematical
Problems in Engineering 2015:1–10. https://doi.org/10.1155/2015/584954.
Mackworth, N. H. 1965. “Visual Noise Causes Tunnel Vision.” Psychonomic
Science 3 (1): 67–68. https://doi.org/10.3758/BF03343023.
Mahapatra, G., A. K. Maurya, and P. Chakroborty. 2018. “Parametric Study of
Microscopic Two-Dimensional Traffic Flow Models: A Literature Review.”
Canadian Journal of Civil Engineering 45 (11): 909–921. https://doi.org/10.
1139/cjce-2017-0686.
Mathew, T. V., C. R. Munigety, and A. Bajpai. 2013. “Strip-Based Approach for
the Simulation of Mixed Traffic Conditions.” Journal of Computing in Civil
Engineering 29 (5). https://doi.org/10.1061/(ASCE)CP.1943-5487.0000378.
Maurya, A. K. 2007. Development of a comprehensive microscopic model for
simulation of large uninterrupted traffic streams without lane discipline. PhD
thesis, Kanpur: Indian Institute of Technology.
Moridpour, S., M. Sarvi, and G. Rose. 2010. “Lane Changing Models: A Critical
Review.” Transportation Letters 2 (3): 157–173. https://doi.org/10.3328/TL.
2010.02.03.157-173.
Munigety, C. R., and T. V. Mathew. 2016. “Towards Behavioral Modeling of
Drivers in Mixed Traffic Conditions.” Transportation in Developing
Economies 2 (1): 6. https://doi.org/10.1007/s40890-016-0012-y.
Newell, G. F. 1961. “Nonlinear Effects in the Dynamics of Car Following.”
Operations Research 9 (2): 209–229. https://doi.org/10.1287/opre.9.2.209.
Nirmale, S. K. 2022. Multi-Vehicle Anticipation-based Models for Describing
Driver Behaviour in Heterogeneous and Disorderly Traffic Conditions. PhD
thesis, Indian Institute of Science.
Nirmale, S. K., and A. R. Pinjari. 2023. “Discrete Choice Models with
Multiplicative Stochasticity in Choice Environment Variables: Application
for Accommodating Perception Errors in Driver Behavior Models.”
Transportation Research Part B: Methodological 170:169–193. https://doi.
org/10.1016/j.trb.2023.02.014.
Nirmale, S. K., A. R. Pinjari, and A. Sharma. 2021. “A Discrete-Continuous
Multi-Vehicle Anticipation Model of Driving Behaviour in Heterogeneous
Disordered Traffic Conditions.” Transportation Research Part C: Emerging
Technologies 128:103144. https://doi.org/10.1016/j.trc.2021.103144.
Nirmale, S. K., A. R. Pinjari, and A. Sharma. 2022. “A Panel Data-Based
Discrete-Continuous Modelling Framework to Analyse Longitudinal Driver
Behavior in Homogeneous and Heterogeneous Disordered Traffic
Conditions.” Transportation Letters 1–14. https://doi.org/10.1080/19427867.
2022.2132058.
Peng, G., and D. Sun. 2010. “A Dynamical Model of Car-Following with the
Consideration of the Multiple Information of Preceding Cars.” Physics
Letters A 374 (15–16): 1694–1698. https://doi.org/10.1016/j.physleta.2010.
02.020 .
Punzo, V., M. Montanino, and B. Ciuffo. 2014. “Do We Really Need to Calibrate All
the Parameters? Variance-Based Sensitivity Analysis to Simplify Microscopic
Traffic Flow Models.” IEEE Transactions on Intelligent Transportation Systems
16 (1): 184–193. https://doi.org/10.1109/TITS.2014.2331453.
Rogé, J., T. Pébayle, E. Lambilliotte, F. Spitzenstetter, D. Giselbrecht, and
A. Muzet. 2004. “Influence of Age, Speed and Duration of Monotonous
Driving Task in Traffic on the Driver’s Useful Visual Field.” Vision
Research 44 (23): 2737–2744. https://doi.org/10.1016/j.visres.2004.05.026.
Saifuzzaman, M., and Z. Zheng. 2014. “Incorporating Human-Factors in
Car-Following Models: A Review of Recent Developments and Research
Needs.” Transportation Research Part C: Emerging Technologies 48:379–403.
https://doi.org/10.1016/j.trc.2014.09.008.
Sarkar, N. C., A. Bhaskar, Z. Zheng, and M. P. Miska. 2020. “Microscopic
Modelling of Area-Based Heterogeneous Traffic Flow: Area Selection and
Vehicle Movement.” Transportation Research Part C: Emerging Technologies
111:373–396. https://doi.org/10.1016/j.trc.2019.12.013.

Sharma, A., Y. Ali, M. Saifuzzaman, Z. Zheng, and M. M. Haque. 2017. “Human
Factors in Modelling Mixed Traffic of Traditional, Connected, and
Automated Vehicles.” International Conference on Applied Human Factors
and Ergonomics 262–273.
Shiomi, Y., T. Hanamori, N. Uno, and H. Shimamoto. 2012. “Modeling Traffic
Flow Dominated by Motorcycles Based on Discrete Choice Approach.”
Proceedings of 1st LATSIS Conference.
Talebpour, A., H. S. Mahmassani, and S. H. Hamdar. 2015. “Modeling
Lane-Changing Behavior in a Connected Environment: A Game Theory
Approach.” Transportation Research Part C: Emerging Technologies
59:216–232. https://doi.org/10.1016/j.trc.2015.07.007.
Toledo, T. 2003. Integrated driving behavior modeling. PhD thesis,
Massachusetts Institute of Technology.
Toledo, T. 2007. “Driving Behaviour: Models and Challenges.” Transport
Reviews 27 (1): 65–84. https://doi.org/10.1080/01441640600823940 .
Treiber, M., A. Hennecke, and D. Helbing. 2000. “Congested Traffic States in
Empirical Observations and Microscopic Simulations.” Physical Review E
62 (2): 1805–1824. https://doi.org/10.1103/PhysRevE.62.1805.
Treiber, M., and A. Kesting. 2013. Traffic Flow Dynamics. Traffic Flow
Dynamics: Data, Models and Simulation. Springer-Verlag Berlin
Heidelberg. https://doi.org/10.1007/978-3-642-32460-4.
Treiber, M., A. Kesting, and D. Helbing. 2006. “Delays, Inaccuracies and
Anticipation in Microscopic Traffic Models.” Physica A: Statistical
Mechanics and Its Applications 360 (1): 71–88. https://doi.org/10.1016/j.
physa.2005.05.001.
TRANSPORTATION LETTERS 19
Walker, J. 2001. Extended discrete choice models: integrated framework, flexible error
structures, and latent variables. PhD thesis, Massachusetts Institute of Technology.
Wang, T., Z. Y. Gao, and X. M. Zhao. 2006. “Multiple Velocity Difference Model
and Its Stability Analysis.” Acta Physica Sinica 55 (2): 634. https://doi.org/10.
7498/aps.55.634.
WHO. (2018). Global Status Report on Road Safety 2018: Summary.
Wilson, R., P. Berg, S. Hooper, and G. Lunt. 2004. “Many-Neighbour Interaction
and Non-Locality in Traffic Models.” The European Physical Journal
B-Condensed Matter and Complex Systems 39 (3): 397–408. https://doi.org/
10.1140/epjb/e2004-00205-y.
Yang, Q., and H. N. Koutsopoulos. 1996. “A Microscopic Traffic Simulator for
Evaluation of Dynamic Traffic Management Systems.” Transportation
Research Part C: Emerging Technologies 4 (3): 113–129. https://doi.org/10.
1016/S0968-090X(96)00006-X.
Yu, L., Z. Shi, and B. Zhou. 2008. “Kink–Antikink Density Wave of an Extended
Car-Following Model in a Cooperative Driving System.” Communications in
Nonlinear Science and Numerical Simulation 13 (10): 2167–2176. https://doi.
org/10.1016/j.cnsns.2007.07.008.
Zhang, X. 2014. “Empirical Analysis of a Generalised Linear Multianticipative
Car-Following Model in Congested Traffic Conditions.” Journal of
Transportation Engineering 140 (6): 04014018. https://doi.org/10.1061/
(ASCE)TE.1943-5436.0000667.
Zheng, Z. 2014. “Recent Developments and Research Needs in Modeling Lane
Changing.” Transportation Research Part B: Methodological 60:16–32.
https://doi.org/10.1016/j.trb.2013.11.009.
20 S. K. NIRMALE ET AL.

Appendix A. Notations

MVA Multiple-vehicle anticipation

αj , βj , γj , α1 , α2 , α3 , λ0 , λ1 , λ2 , pj ; qj Sensitivity or weighting coefficients
V1 , V2 , C, C0 C1 , and C2 Constants
λ Parameter
ai ðtÞ Acceleration/deceleration of subject vehicle i at time t
amax
i Maximum acceleration of vehicle i
b Braking deceleration
bconf
i
Comfortable deceleration of vehicle i
M Mass of the vehicle
N, N1 , N2 Number of lead vehicles
i Subject vehicle index
j Lead vehicle index
t Time instant
Vi ðtÞ Velocity of vehicle i at time instant t (m/s)
Vequ Velocity at equilibrium
ðdesÞ
Vi ðtÞ Desired velocity for subject vehicle iat time t
Vð:Þ Optimal velocity function
V0 Free speed
Ve Equilibrium speed spacing function
Vmax Maximum speed
ðjÞ
ðjÞ
ΔVi ðtÞ speed difference between the speed of the jth lead vehicle and speed of the subject vehicle i, i.e. ΔVi ðtÞ ¼ Vj ðtÞ Vi ðtÞ
Xi ðtÞ Position of vehicle i at time instant t
X0 Stopping distance, including vehicle length (m)
ΔXc Safe distance which is considered as constant
ðjÞ
ΔXi ðtÞ Distance Xj ðtÞ Xi ðtÞ (m) between lead vehicle j and subject vehicle i

ΔXi
ðj;desÞ
ðtÞ Desired following distance (m) for subject vehicle i at time t with respect to lead vehicle j
ðjamÞ
ΔXi ðtÞ Minimum spacing at the standstill situation
ðminÞ
ΔXi Minimum space headway for subject vehicle i
ðsafeÞ
ΔXi ðtÞ Safe distance
ΔXequ Space gap at equilibrium
Li Length of vehicle i
LSi
ðjÞ Lateral separation between subject vehicle i and lead vehicle j
LSmax Maximum lateral separation
ðjÞ
Δai ðtÞ Acceleration difference between lead vehicle j and subject vehicle i at time t
ðcomf Þ
di Comfortable or desired deceleration
Ti Reaction time or delay time of subject vehicle i
Ti
ðdesÞ Desired time headway of vehicle i
TGap Time gap
Δt Update time
ðjÞ
ðdesÞ
Vi ðΔXi ðtÞÞ
ðjÞ Desired speed of subject vehicle i for a given space headwayΔXi ðtÞat time t
Uij Utility that the driver of the ith vehicle perceives from choosing a maneuvering decision j
Zij ; Kij Traffic environment variables (such as space gap, relative speed) with respect to the driver of the ith vehicle and maneuvering
decision j
Senð:Þ Sensitivity function
Stimð:Þ Stimulus function