J Control
Theo
η
Appl
2012
10
(4)
426
--4
34
DOI
1O
.1007/s 11768-012-0319-0
Vehicle adaptive cruise control design with optimal
switching between throttle
and
brake
Lihua
LUO
1, Ping LI 2, Hui WANG 2
l.
College
ofTransport
and
Communication
,
Shanghai
Maritime
University
,
Shanghai
200135
,
China;
2.Stat
巳
Key
Laboratory
o
fI
ndustrial
Control
Technology
,
Zhejiang
University
,
Hangzhou
Zhejiang
310027
,
China
Abstract: For vehicle adaptive cruise control
(AC
C) systems,
the
switching performance between throttle
and
brake
determines
the
driving comfort,
fuel
consumption
and
servic
巳
lives
of vehicle
m
巳
chanical
compon
巳
nts.
In this paper,
an
ACC
a1
gorithrn with the optim
a1
switching control between throttle
and
brake
is
designed
in
model predictive control
(MP
C)
framewor
k.
By
introducing the binary integer variables, the dynamics of throttle
and
brake
are
integrated
in
on
巳
model expression
for
the
controller design. Then the
ACC
algorithm
is
designed
to
satisfy not only safe car following,
but
also
the
optim
a1
switching between throttle
and
brake, which leads
to
an
online mixed integer quadratic programming solved
by
the
nested two-loop method. The simulation results
show
that
由
e
proposed
ACC
algorithrn meets
the
requirements of
safe car following
, outperforms
the
traditional algorithms
by
performing smoother responses, reducing
the
switching
times
between throttle
and
brake,
and
th
巳
refore
improves driving comfort
and
fuel
efficiency significantly.
Keywords: Adaptive cruise control; Model predictive control; Optimal switching;
Throttle; Brake
1 Introduction
Advanced driver assistant (ADA)
systems
缸
e
in
位
0-
duced to relieve human drivers from routine tasks, increase
drivers' safety and comfort
, reduce fuel consumption, and
improve freeway capacity. Examples
of
ADA
systems
缸
e
adaptive cruise control (ACC) systems, Lane-Keeping sys-
tems, collision avoidance (CA)
syst
巳
ms
,
and etc.
[1
-3].
An ACC vehi
c1
e adjusts its speed at a
p
时
efined
value
when
there is no preceding vehi
c1
e. Wh
en
a preceding ve-
hi
c1
e is detected
in
the same lane, it automatically follows
the preceding vehi
c1
e with a desired spacing. The control
architecture for an ACC system is usua
l1
y designed to
be
hi
巳
rarchical
and consists
of
two controllers [3]:
th
巳
upper
level contro
l1
er
det
巳
rmines
the desired acceleration to sat-
isfy the requirements based on measured sign
a1
s; and the
lower level controller activates the corresponding actuator
(throttle
or
brake) according to the desired acceleration from
the
upper
level controller. In the lower level controller, the
switching strategy between throttle and brake is tradition-
ally
based
on the simple threshold criteria
[4-
6]:
if
the de-
sired acceleration from the upper level controller is larger
than a critic
a1
v
a1
ue
, the throttle is activated,
or
the brake is
activated.
While
in
[7], the switching rule is determined ac-
cording to the current traffic situation
, which will cause that
the switching is sensitive to the surrounding traffic condi-
tion. Moreover
, researchers have proposed a switching cri-
terion by
ca
1c
ulating the throttle and brake laws simulta-
neously and judging the corresponding sign
of
two control
laws
[8]. Above all, it can be seen that the switching con-
trol between
throttl
巳
and
brake is not fully considered in
the procedure
of
ACC algorithm design, which may cause
仕巳
quent
switching, and thus lead to uncomfortable driving,
increased fuel consumption and mechanical abrasion
of
ve-
hi
c1
e components.
Received
27
Demceber
2010;
revised
19
August
2011
In this paper, an ACC algorithm with the optimal switch-
ing between throttle and brake is designed
in
model pre-
dictive control
(MP
C)仕
amewor
k.
Song's
longitudin
a1
ve-
hi
c1
e model [9] is considered for the design. First, the dy-
narnics
of
throttle and brake, which cannot be activated si-
multaneously
,缸巳
integrated
in
one state-equation model
by
introducing binary integer variables.
Then
the goal
of
minimizing the switching between throttle and brake, to-
gether with the safe car-following requirements and vehi-
c1
e capabilities are modeled as the control objectives and
constraints
of
ACC, respectively.
The
contro
l1
er design is
transformed to be an online mixed integer quadratic pro-
gramming (MIQP)
, which is solved by the developed nested
two-loop method.
By
simulating five representative traffic
scenarios
, it is
proved
出
at
the proposed ACC algorithm
provides safe vehi
c1
e
followi
吨,
and meanwhile reduces
switching between throttle and brake
, therefore improves
driving comfort and fuel efficiency compared to the tradi-
tional algorithm.
The reminder
of
this paper is organized as follows: Sec-
tion 2 is the modeling
of
ACC system,
in
which the bi-
nary integer variables are introduced to integrate the throttle
and brake dynamics. Section 3 represents the contro
l1
er
de-
sign
of
ACC, in which the switching between throttle and
brake is
fu
l1
y considered and optimized. In Section 4, the
proposed optimal switching algorithm and the traditional
threshold switching algorithm
ar
巳
simulated
and analyzed
in various traffic situations. The con
c1
usion is represented
in the Section 5.
2 ACC system modeling
For ACC vehi
c1
e, Song' longitudinal vehi
c1
e model [9] is
considered
, which assumes that the inertia
of
the engine and
This
work
was
suppo
口
ed
by
Science
&
T
,巳
chnology
Program
of
Shanghai
Maritime
University
(No.
20120077)
@
South
China
University
ofTechnology
and
Academy
ofMathematics
and
Systems
Science
,
CAS
and
Springer-Verlag
Berlin
Heidelberg
2012