Proceedings of the 3rd International Conference on TuCC.

2
Systems and Control, Algiers, Algeria,
October 29-31, 2013

Software in the loop model and decoupling control for dual clutch
automotive transmissions
Maurizio Zoppi\ , Claudio Cervone♦ , Gerardo Tiso♦ , Francesco Vasca\

Abstract— Dry dual clutches are widely used in conventional controllers, particularly for dual clutch systems, are dif-
and innovative automotive vehicles. The engagement control of ficult to be designed without an accurate model of the
dry dual clutches is a not easy task because of the synergic torque transmitted by the clutches, and more specifically,
interaction of the two clutches in terms of their corresponding
torque transmissibility characteristics. In this scenario the so- of the relationship between the clutch actuator position
called software in the loop approach can represent a founda- (or the pressure applied by the clutch actuator) and the
mental preliminary and driving step for hardware in the loop torque transmitted through the clutch during the engagement
and on vehicle testing of new engagement strategies. This paper phase [12], [13], [14], [15], [16], [17]. The software in
proposes a software in the loop model for dry dual clutch the loop DCT model proposed in this paper is based on
transmission control systems, developed by Fiat for new C635
transmission family and its future evolutions. The vehicle model a real automotive DCT system by Fiat and consists of
includes the representation of the shift process for a six-speed two main subsystems: a transmission dynamic model and a
dry-type dual clutch transmission. The transmission model is transmission control unit (TCU). The former includes models
used to design a decoupling engagement controller of the dual for engine, gearbox, pump and clutch actuators for odd and
clutch. even sides. The main TCU task consists of the management
Keywords: Automotive control, dry clutch, dual clutch, of upshift and downshift of the dual clutch system, e.g. to
driveline control, clutch engagement, software in the loop. compute and to send the control signals to the actuators based
on the driveline measurements [18]. The development of the
I. I NTRODUCTION TCU model is not only dedicated to the model validation but
Modern dual clutch transmissions (DCTs) combine the also to the possibility of having a simple and intuitive off-line
advantages of fully automated and automated manual trans- system for control testing [19], [20]. The effectiveness of the
missions [1]. Indeed DCT, differently from single clutch proposed software in the loop system is shown by testing a
systems, are capable of shifting without interruption of the new DCT regulation strategy based on the decoupling control
traction power [2], [3]. An advantage with respect to so- technique.
called wet solutions is in terms of efficiency since there is no
II. M ODELS OF SUBSYSTEMS
need to provide the oil volume flow that would otherwise be
required for cooling purposes. Moreover, differently from a The proposed software in the loop transmission model
fully automatic transmission, DCTs don’t require torque con- is based on the C635 dry DCT system shown in Fig. 1.
verters. Overall the essential advantage of DCT technology is In order to motivate the model decomposition into the
the combination of efficiency and driving enjoyment [4], [5]. different subsystems it is important to describe first the
A two-part transmission shaft is at the heart of a DCT. main operations and elements of a DCT. During the shifting
The DCT splits up odd and even gears on two input shafts. process the transmitted torque is obtained by the overlap
The outer shaft is hollowed out, making room for an inner of the engagement of the closing clutch and the release
shaft, which is nested inside. The outer hollow shaft feeds of the opening clutch. The clutch K1 is normally closed
even gears, while the inner shaft feeds odd gears [6]. As as in conventional manual transmissions and its position is
the vehicle moves through one gear - second, for example controlled by means of a contact-less linear position sensor
- the other gearbox prepares the next expected gear - in integrated in the rear hydraulic piston actuator. The even gear
this case third - so that the change can be completed in a clutch K2 is normally open and is controlled in force, i.e.,
smooth, precisely timed process [7]. Than one clutch controls through hydraulic pressure. The two clutches act on a center
even gears, while the other clutch controls odd gears, that’s plate together with the two pressure plates. The entire dual
the feature that allows lightning-fast gear changes and keeps clutch unit is mounted on the clutch housing by means of a
power delivery constant [8]. single main support bearing. This compact mounting solution
The clutch control design and performance analysis phases was developed thanks to the adoption of the specific actuation
can be highly supported by the use of software and hardware system of the clutch K1 which allows the space for such a
in the loop technologies [9], [10], [11]. Indeed effective bearing to be installed.
An equivalent scheme for the dynamic model of the
(\) Department of Engineering, University of Sannio, Piazza Roma 21, powertrain with the dry DCT is shown in Fig. 2. The
82100 Benevento, Italy, email: {mzoppi, vasca}@unisannio.it scheme includes an engine, a dry dual clutch, a gearbox
(♦) FGA, EMEA Region, Powertrain Engineering, Controls,
Pomigliano Technical Center, Napoli, Italy, email: {Claudio.CERVONE, and a final reducer, as well as equivalent inertias modeling
gerardo.tiso}@fiat.com wheels and vehicle. The first clutch (K1) is used for odd

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on TuCC.2

where ωin and Jin are the speed and the inertia of the
flywheel, respectively. The variables TK1 and TK2 represent
the torques transmitted by the two clutches and Tin is the
flywheel torque, which can be expressed as
Tin = B1 (θe − θin ) − C1 (ωe − ωin ) (4)
where B1 and C1 are the torsional rigidity and the damping
Fig. 1. Dry dual clutch unit. coefficient of the input shaft, respectively. The variables θe
and θin are the rotational angles of the engine crankshaft and
the flywheel:
gears (1st, 3rd and 5th) while the second clutch (K2) is
used for the even ones (2nd, 4th and 6th). Both clutches θ̇e = ωe (5a)
are connected to the input shaft. When a gear shift signal θ̇in = ωin . (5b)
is initiated, the opening clutch is released and the closing
clutch simultaneously engaged, which allows torque transfer The other dynamic equations of the different nodes of the
without power interrupted during a gearshift. As shown in driveline can be expressed as
Fig. 2, in the software in the loop model all gears and shafts JK1 ω̇K1 = TK1 − Tgodd (6a)
in the gearbox are simplified with equivalent inertias and
torsional elastic-damper elements, and the two clutches are JK2 ω̇K2 = TK2 − Tgeven (6b)
characterized by friction elements with inertias. Jodd ω̇odd = igodd Tgodd − Todd (6c)
Jeven ω̇even = igeven Tgeven − Teven (6d)
K1
JK1
Tgodd Jout ω̇out = if Todd + if Teven − Tout (6e)
B1 ωK1
igodd
where JK1 , JK2 , Jodd , Jeven and Jout are the inertias of
Engine Jin Jodd
ωe ωin ωodd
Todd
B2 K1, K2, odd gear, even gear and output shaft assemblies,
C1
K2 if respectively, ωK1 , ωK2 , ωodd , ωeven and ωout are the speeds
Ce
JK2
ωK2
Tgeven
Jout
ωout
V ehicle
ωw
of K1, K2, odd gear, even gear and output shaft, respectively,
igeven if igodd and igeven are the gear ratios of the odd and even gears,
C2
Jeven Teven
if is the final drive gear ratio, Tout is the torque at the output
ωeven
shaft, and Tgodd , Tgeven , Todd and Teven are the torques at
Fig. 2. Equivalent scheme of the driveline with dry dual clutch. the different nodes of the scheme in Fig. 2.
The equilibrium equation for the output shaft can be
In what follows the models used for the different subsys- expressed in the following form
tems are described.
Tout = B2 (θout − θw ) − C2 (ωout − ωw ) (7)
A. Engine where θout and θw are the rotational angles of the output
As usual for automotive driveline control problems, the shaft and the wheels, ωout and ωw are the speeds of the
engine torque is modeled by means of a static map depending output shaft and the wheels, B2 is the stiffness of the output
on the throttle angle and on the engine speed: shaft, and C2 is the damping coefficient of the output shaft.
The rotational angles θout and θw satisfy the following
Te = f (αe , ωe ) (1)
equations
where αe is the throttle angle and ωe is the engine speed.
θ̇out = ωout (8a)
According to the equivalent scheme reported in Fig. 2, the
time evolution of the engine crankshaft speed is regulated by θ̇w = ωw . (8b)
the following dynamic equation By considering the relationship between the longitudinal
Je ω̇e = Te − Tf − Tin (2) vehicle velocity and the wheels speed, i.e., v = Rw ωw , the
wheels rotation dynamics can be written as
where Je is the inertia of the engine crankshaft, Tf is the
torque due to the friction losses and Tin is the flywheel Jw ω̇w = Tout − Tw (9)
torque wich becomes the input for the rest of the driveline with Jw = mRw 2
and Tw = F0 + F1 v 2 , where F0 and F1
model. are appropriate coefficients.
B. Dual clutch transmission driveline The driveline model in the classical state-space form can be
obtained by assuming ideal gears and then
The equilibrium equation of motion in torsional vibrations
for the input shaft of the dry dual clutch can be written as ωK1 = igodd ωodd = igodd if ωout (10a)
Jin ω̇in = Tin − (TK1 + TK2 ) (3) ωK2 = igeven ωeven = igeven if ωout . (10b)

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on TuCC.2

By substituting (10) into (6) and by using simple algebraic together with
manipulations it is possible to write the following dynamic TK1 = 0. (20)
equation
Therefore the model (13) can be used in all DCT operating
JDCT ω̇out = igodd if TK1 + igeven if TK2 − Tout (11) conditions. During the engagement Te , TK1 and TK2 are
where obtained from the engagement controller. When K1 (K2) is
locked-up and K2 (K1) is fully open the clutch torques are
JDCT = i2godd i2f JK1 + i2f Jodd + i2geven i2f JK2 + i2f Jeven . given by (18) and (15b) ((20) and (19), respectively) while
(12) the engine torque responds to the driver commands. The
Then the dynamic model of the driveline is given by (1) clutches speeds can be obtained by using (10) with ωout =
− (5), (11) − (12) and (7) − (9). This model is sufficiently ωw . The reduced order model (13) has been implemented and
detailed in order to be implemented into the software in the used for the realization of the engagement control design.
loop setup, but it is too complex for the dry dual clutch
control design. A simplification of the model presented above C. Actuation module
can be obtained by assuming the rigidity of the input and The clutches and gear shifting mechanisms of the C635
transmission output shafts. In this case the speeds ωout and DCT are electro-hydraulically actuated through a dedicated,
ωw can be considered to be equal, as well as ωe and ωin . sealed, hydraulic oil circuit, see Fig. 3. The system is com-
Then the simplified dynamic model of the driveline can be posed of a hydraulic power unit, consisting of an electrically
written as driven high pressure pump and accumulator, and an actuation
Tef − (TK1 + TK2 ) module which includes the control solenoid valves, gear shift
ω̇e = (13a)
Jein actuators and sensors [21].
igodd if TK1 + igeven if TK2 − Tw The clutch and gear actuation module consists of 4 distinct
ω̇w = (13b) double action pistons operating the gear engagement forks
JDCT w
(lower right part in Fig. 3), one shifter spool which selects
where
the piston to be actuated and 5 solenoid valves of which 4
Jein = Je + Jin (14a) are pressure proportional (PPV) and one flow proportional
Tef = Te − Tf (14b) (QPV). PPV1 and PPV2 actuate the gear engagement piston
which is selected by the spool valve operated PPVS. The
JDCT w = JDCT + Jw . (14c)
valve PPV-K2 is used for the control of the concentric slave
The model (13) is valid not only during the engagement cylinder of the clutch K2, while QPV-K1 is used for the
phases, when the clutches torques are determined by the position control of the clutch K1. The primary shaft of the
corresponding actuators, but also when one of the the two gearbox is composed of two parts. On one of these two parts
clutch is locked-up and the other is fully open. For instance are all the odd gears (1st, 3rd, 5th, Rm) and is connected
consider the case when the clutch K1 is locked-up and the to the clutch K1, on the other hand we find all the even
clutch K2 is fully open. Then one can write gears (2nd, 4th, 6th) and is connected to the clutch K2.
The selection of plans graft (rank) is handled through the
ω̇e = ω̇K1 (15a) proportional pressure solenoid valve PPV-S which controls
TK2 = 0. (15b) the spool (shifter) of hydraulic power.
By using (10a) with ωw = ωout it follows
K2
ωK1
ωw = (16) P P V -K2
igodd if
K1
and substituting (16) together with (15b) into the rela- QP V -K1

tion (13b) one obtains

i2godd i2f TK1 − Tw igodd if 2 1
Shif ter

ω̇K1 = . (17) PPV 2

JDCT w A1
1 < −− > 3
By using (13a) and (17) in (15) one can write PPV 1
A2
4 < −− > 2
Tef JDCT w + Tw Jein igodd if
TK1 = (18) A3
JDCT w + Jein i2godd i2f P P V -S
5 < −− > RM

A4
which provides the torque transmitted by K1 when that clutch U.I. 6 < −− > (7)

is locked-up. By following a similar procedure the torque

transmitted by K2 when it is locked-up and K1 is fully open Fig. 3. Equivalent scheme for the electrohydraulic valves regulation.
can be written as
Tef JDCT w + Tw Jein igeven if As an example, by looking at Fig. 3, consider a possible
TK2 = (19)
JDCT w + Jein i2geven i2f sequence of the engagements of the 1st gear and then of

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on TuCC.2

the 2nd gear. The TCU controls the PPV-S in the first
position and, at the same time, the solenoid valve PPV2
sends pressurized oil along the channel colored red. The
spool valve positioning itself in the first position, puts in
communication the red channel with the left side of the
actuator A1. At this point the oil fills the left chamber of the
actuator by moving the latter to the right engaging the 1st
gear. Obviously this operation is performed after the control
unit TCU via the solenoid valve QPV-K1 has proceeded to
open the clutch K1 commanding the relative actuator. In
the moment in which it is engaged in 1st gear, the control
unit de-energizes the QPV-K1 so as to drain the oil and to
reclose the clutch K1. During the transmission of motion
in 1st gear, there is the pre-selection of the 2nd gear. The
control unit TCU controls the spool through the PPV-S in
order to place it in the second position, at the same time the
Fig. 4. Transmissibility clutch torque vs. the actuator position at constant
control unit TCU controls the solenoid valve PPV1 in order temperature for a given wear.
to be loaded with oil under pressure the blue color channel.
The spool in second position puts in communication the blue
channel with the left side of the actuator A2. In this way, the
pressurized oil pushes the actuator A2 to the right engaging
By substituting (22) in (13) one can write
the 2nd gear. All the operations described above and those
corresponding to the other gearshifts, are implemented into
corresponding tasks of the TCU model, while valves and Tef − (TK1 + TK2 )
pistons are modeled by means of static maps and second ω̇e = (23a)
Jein
order dynamic systems. ig(k) ig(k+1) i2f TK1 + i2g(k+1) i2f TK2 − ig(k+1) if Tw
D. Transmissibility models ω̇K2 = .
JDCT w
The two clutches act on a center plate which, together with (23b)
the two pressure plates, has been dimensioned according to
the thermal dissipation characteristics required for the most By using the model (23) it is possible to design a control
critical vehicle-engine applications foreseen. The actuator strategy which decouples the regulations of the engine and
positions determine the corresponding torque transmitted closing clutch speeds. Let us define the following fictitious
by the clutches during the engagement and disengagement control signals:
phases through the clutch transmissibility characteristics,
see [12]. This transmissibility models are also used in the Te − TK1 − TK2
control algorithms of the TCU and form the basis of the en- u1 = (24a)
Jein
gagement and thermal protection strategies. Fig. 4 represents
ig(k) ig(k+1) i2f TK1 + i2g(k+1) i2f TK2
a typical clutch torque transmissibility characteristic. u2 = . (24b)
JDCT w
III. E NGAGEMENT CONTROL DESIGN
The clutch control strategy depends on whether the clutch By substituting (24) in (23) one obtains
is opening or closing. In particular for the clutch to be
opened, say K1, an open-loop control is adopted, which at
Tf
appropriate time opens the clutch with a given slow rate of ω̇e = u1 − (25a)
the actuator variable. Instead, the control of the clutch to be Jein
closed, say K2, is in closed-loop. Define ig(k) and ig(k+1) ig(k+1) if Tw
ω̇K2 = u2 − (25b)
the gear ratios at the beginning and at the end of the shift- JDCT w
phase respectively. Without loss of generality it is assumed
which is a decoupled two-input (u1 and u2 ) and two-output
that K1 is opening and K2 is closing and then
(ωe and ωK2 ) system where Tf and Tw act as disturbances.
ig(k) = igodd (21a) Then the models of the two decoupled subsystems are simple
ig(k+1) = igeven . (21b) first order integrators and the control signals u1 and u2 can be
obtained by using standard proportional-integral controllers
The relation between the wheels and the closing clutch on the speeds errors with respect to the reference values of
speeds is given by ωe and ωK2 , respectively. The feedback control provides u1
ωK2 and u2 from which the reference values for the engine and
ωw = . (22)
ig(k+1) if closing clutch torques must be obtained. By inverting (24)

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on TuCC.2

one can write suitable conditions for shift are achieved. The engagement
JDCT w ig(k)   controller is turned off when the positions of the two
Teref = Jein u1 + u2 + 1 − TK1 (26a) actuators reach their end-limits.
i2g(k+1) i2f ig(k+1)
ref JDCT w ig(k) IV. S OFTWARE IN THE LOOP SIMULATIONS
TK2 = 2 u2 − TK1 (26b)
ig(k+1) i2f ig(k+1) The software in the loop model of the DCT with the
TCU has been developed and the decoupling controller
where TK1 is given by the open-loop strategy for the opening
implemented in the Matlab/Simulink environment. The trans-
clutch K1.
mission system controls the shift schedule based on the
Fig. 5 shows the scheme with the decoupling control,
vehicle speed and throttle angle. Before the shift the clutch
where the inversion block of the transmissibility character-
to be closed is connected to the driveline with the new
istic of the two clutches is used in order to determine the
gear ratio ig(k+1) . In the successive torque phase the engine
two actuators signals corresponding to the reference torques.
torque is transferred by both clutches simultaneously and the
The constant engine speed reference ωeref is appropriately
two clutches are partially closed. This is the most critical
calculated trying to make the shift as much as possible
phase for dry DCT gearshifts and the engagement control
comfortable. A good gear shift quality can be achieved by
action plays a key role in order to ensure a constant power
maintaining a constant vehicle speed. Indicate with v(k) and
transmission and to avoid driver discomfort. Fig. 6 shows the
v(k + 1) the vehicle speeds at the beginning and at the end
engine and clutches speeds transmission obtained by using
of the shift, respectively:
the decoupling engagement control presented in the previous
section. Note that the warming phase ends at 5s and the
Rw duration of the engagement phase at start-up is about 3s.
v(k) = ωe (k) (27a)
ig(k) if
Rw
v(k + 1) = ωe (k + 1). (27b)
ig(k+1) if
The engine speed reference is then obtained by assuming
v(k) = v(k + 1) in (27) which provides
ig(k+1)
ωeref = ωe (k + 1) = ωe (k). (28)
ig(k)

ωe Tf

+ − u1
ωeref P Ie + Teref

δ1ref
ref
TK1
f (δ, T ) TK1 (δ1ref )
∗
ig(k)
ig(k+1)
Fig. 6. Clutches, engine and wheels speeds for an upshift sequence with
the decoupling engagement controller.

Fig. 7 shows the clutch torques during gear changes.

ref + u2 During the gear change there are time intervals in which both
ωK2 P Ic −
−
1
∗ + δ2ref (TK2 ) δ2ref clutches torques contribute to the motion of the vehicle. After
i2g(k+1) i2f the engagement phase the torque transmitted by the closed
ωK2
Tw
ig(k+1) if
clutch is equal to the engine torque.
So as described in Sec. II, the software in the loop model
Fig. 5. Decoupling control scheme. includes also the DCT actuation model. When the clutch
should open or close the transmission control unit sends
Coherently with (27), by using (22) also the reference current signals to the actuators that allow the opening or
signal for the clutch speed is chosen equal to (28). closing of the clutch. In the upper part of Fig. 8 it is shown
At vehicle start-up the first gear is already engaged and for the first actuator position. When the position of the first
the engagement of the clutch K1 the engine speed reference actuator assumes its negative end-limit value the 1st gear
is chosen ad-hoc, so that the departure has a proper duration is engaged, instead, when it assumes its positive end-limit
and a good comfort is ensured. value the 3rd gear is engaged. The lower side of the Fig. 8
The TCU model activates the dual clutch control when shows the pressure of the clutch K2. The upper constant
there is a gear change request from the driver or when levels of the pressure correspond to the even gears engaged.

978-1-4799-0275-0/13/$31.00 ©2013 IEEE

Proceedings of the 3rd International Conference on TuCC.2

[2] M. Kulkarni, T. Shim, and Y. Zhang, “Shift dynamics and control of

dual-clutch transmissions,” Mechanism and Machine Theory, vol. 42,
pp. 168–182, 2006.
[3] J. Zhang, M. Wu, and T. Lu, “Optimal engagement of dry dct vehicle
launch by feedback linearization control,” in Proc. of the International
Symposium on Advanced Vehicle Control, Loughborough, UK, 2010,
pp. 930–935.
[4] A. S. Joshi, N. P. Shah, and D. C. Mi, “Modeling and simulation of
a dual clutch hybrid vehicle powertrain,” in Proc. of the IEEE Vehicle
Power and Propulsion Conference, Dearborn, Michigan, USA, 2009,
pp. 1071–1078.
[5] Y. Liu, D. Qin, H. Jiang, and C. Liu, “Shift schedule optimization
for dual clutch transmissions,” in Vehicle Power and Propulsion
Conference, Dearborn, Michigan, USA, 2009, pp. 1666–1673.
[6] E. Galvagno, M. Velardocchia, and A. Vigliani, “Dynamic and kine-
matic model of a dual clutch transmission,” Mechanism and Machine
Theory, vol. 46, pp. 794–805, 2011.
[7] I.-G. Jang, I. Seo, J. Jeon, and S.-H. Hwang, “Testing 32-bit embedded
system using hardware-in-the-loop-simulation of automatic transmis-
sion,” in Proc. of the International Conference on Control, Automation
and Systems, Coex, Seoul, Korea, 2007, pp. 399–403.
[8] H. Hu, X. Wang, and Y. Shao, “Optimization of the shift quality
Fig. 7. Clutches torques. of dual clutch transmission using genetic algorithm,” in Proc. of the
International Conference on Natural Computation, Yantai, Shandong,
China, 2010, pp. 4152–4156.
[9] H. Quanan, S. Jian, and L. Lei, “Research on rapid testing platform for
The signal evolution also shows the dynamic variation of the tcu of automated manual transmission,” in Proc. of the International
simulated pressure. Conference on Measuring Technology and Mechatronics Automation,
Shangshai, China, 2011, pp. 67–70.
[10] W. Lhomme, R. Trigui, A. Bouscayro, P. Delarue, B. Jeanneret, and
F. Badin, “Validation of mechanical transmission with clutch using
hardware-in-the-loop simulation,” in Proc. of the IEEE Vehicle Power
and Propulsion Conference, Arlington, Texas, USA, 2007, pp. 425–
431.
[11] L. Mianzo, “A transmission model for hardware-in-the-loop powertrain
control system software development,” in Proc. of the IEEE Interna-
tional Conference on Control Applications, Anchorage, Alaska, 2000,
pp. 1–8.
[12] F. Vasca, L. Iannelli, A. Senatore, and G. Reale, “Torque transmissi-
bility assessment for automotive dry-clutch engagement,” IEEE/ASME
Transactions on Mechatronics, vol. 16, no. 3, pp. 564–573, 2011.
[13] Y. Liu, D. Qin, H. Jiang, C. Liu, and Y. Zhang, “Clutch torque formu-
lation and calibration for dry dual clutch transmissions,” Mechanism
and Machine Theory, vol. 46, pp. 218–227, 2011.
[14] C. Ni, T. Lu, and J. Zhang, “Gearshift control for dry dual-clutch
transmissions,” WSEAS Transactions on Systems, vol. 8, no. 11, pp.
1177–1186, 2009.
[15] D. Reed, I. Kolmanovsky, and D. Hrovat, “Robust control methods
applied to an automated clutch transmission,” in Proc. of the Interna-
tional Symposium on Advanced Vehicle Control, Loughborough, UK,
2010, pp. 663–668.
Fig. 8. Actuator position and clutch K2 pressure. [16] P. D. Walker and N. Zhang, “Engagement and control of synchronizer
mechanisms in dual clutch transmissions,” Mechanical Systems and
Signal Processing, vol. 25, pp. 1923–1936, 2011.
V. C ONCLUSIONS [17] T. Szabo, M. Buchholz, and K. Dietmayer, “Model-predictive control
of powershifts of heavy-duty trucks with dual-clutch transmissions,”
The main elements of a software in the loop model for dry in Decision and Control (CDC), 2012 IEEE 51st Annual Conference
dual clutch transmissions have been presented. The model on, 2012, pp. 4555–4561.
is based on the new C635 transmission family and can be [18] D. Xue, X. Yin, and L. Li, “Software architecture for the ecu
of automated manual transmission,” in Proc. of the International
used for testing clutch engagement control algorithms before Conference on Natural Computation, Yantai, Shandong, China, 2010,
their direct porting on a hardware in the loop platform. A pp. 63–68.
new engagement control strategy for the dual clutch system [19] W. Yang, G. Wu, and J. Dang, “Research and development of
automatic transmission electronic control system,” in Proc. of the IEEE
has been proposed. The controller is obtained by using a International Conference on Integration Technology, Shenzhen, China,
simplified driveline model and by decoupling the regulations 2007, pp. 442–445.
of the engine speed and that of the closing clutch. The [20] J. Tan, X. Yin, and L. Yin, “Can-based integrated control strategy
for the drivetrain of commercial vehicles equipped with automated
software in the loop simulations have shown the effectviness manual transmission,” in Proc. of the Workshop on Power Electronics
of the proposed control strategy. and Intelligent Transportation System, Guangzhou, China, 2008, pp.
67–71.
R EFERENCES [21] C. Vafidis, F. Cimmino, and C. Cervone, “Fpt’s high torque density
dual dry clutch transmission (htd-ddct),” in International CTI Sym-
[1] A. Senatore, “Advances in the automotive systems: An overview of posium Innovative Automotive Transmissions, Berlin, Germany, 2009,
dual-clutch transmissions,” Recent Patents on Mechanical Engineer- pp. 1–13.
ing, vol. 2, no. 2, pp. 93–101, 2009.

978-1-4799-0275-0/13/$31.00 ©2013 IEEE