Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Abstract
In order to make modern Diesel engines cleaner and more fuel efficient, their air systems
architecture become more and more complex. The control strategies of these systems must take
account of the multiple components interactions with minimal calibration effort required. In this
context, model based techniques are very attractive. In this paper we propose a control oriented
model of a variable geometry turbocharger in an architecture with two exhaust gas recirculation
(EGR) loops : high pressure (HP) and low pressure (LP). This model is implemented in a basic
control strategy and evaluated experimentally during tests with LP or HP EGR. The results show
that the choice of EGR circuit has a high influence on the turbocharger actuator position, but that
this effect is well taken into account in the model proposed.
• Low Pressure EGR: Gas are taken downstream the 2.1 Turbocharger modeling
particulate filter and derived to the upstream of the
compressor. On the contrary to the HP EGR, this The turbocharger is composed by a turbine driven by the
dynamics is much slower but the turbocharger exhaust gas and connected via a common shaft to the
takes all the energy of the exhaust gases. compressor, which compresses the air in the intake. The
rotational speed of the turbocharger shaft Nt can be
derived from a balance between the turbine power P t
and the compressor power Pc
Control input The turbocharger is equipped with guide
vanes whose angles are adjusted via an actuator that is d ( 1 J N 2) = P − P (1)
noted uvgt. This affects both the angle of the gas flow on dt 2 t t t c
the turbine blades and the turbine effective flow area. By where Jt is the inertia of the turbocharger.
these means, it is possible to maintain a high boost even Compressor In order to derive an equation for the com-
at low engine speed, and to improve the system pressor power, the first law of thermodynamics is applied. It
dynamics performance. states that (neglecting heat losses) the compressor power
is related to the mass flow through the compressor Dc and
the total change of enthalpy by Pc = Dccp(Tdc −Tut). The
Measurements The sensors available on the system are compressor efficiency is introduced as the ratio between
the following : isentropic and actual compression powers. The
compressor power reads
• Engine speed Ne. Pc = DccpTuc ηc
1
Πc γ
γ−1
−1 (2)
Puc and Tuc. compressor pressure ratio, and γ the specific heat ratio.
• Manifold air flow Dc.
The compressor speed, flow, pressure ratio and efficiency are
The other variables will be estimated with the measured linked. Different representations can be found in the literature,
variables among which a commonly used one consists in mapping the
pressure ratio and efficiency against flow and speed. These
1.3 Model objective maps are extrapolated from data measured during
characterization tests. Several extrapolation meth-ods have
been proposed (for example Jensen et al. [1991]). In order to
First of all, the objective of the model development de- take the variations of the upstream compressor
scribed here is to provide a basis for the design of a conditions into account, these variables are corrected as
control strategy. The difficulty of this task consists in
keeping the right level of complexity. Two main criteria
follow √
T Nt
uc
will be considered : D D P
c,cor = and Nc,cor = √ Tuc
c uc
• The model has to represent only the main dynamics and
governing the evolution of the system in order to ηc = φηc (Dc,cor , Nc,cor ) (3)
minimize the number of states in the controller. The Π = φ (D ,N )
c Πc c,cor c,cor
fast dynamics are neglected.
• In an engine, the evolution of a turbocharger depends The compressor pressure ratio corresponding to the
on the conditions at its boundaries : pressures and system studied here is represented in Figure 2.
temperatures upstream and downstream the com-
pressor and turbine, gas mass flow through these Turbine Similarly, the turbocharger power is related to
com-ponents. This is very important particularly for the the mass flow through the turbine Dt and the total
architecture with two EGR loops that is considered in change of enthalpy. This results in
Pt = DtcpT utηt γ (4)
1 − Πt γ
1−
this paper. These influences must be represented as
far as possible, using measurements when available
or estimations otherwise. where ηt is the turbine efficiency, Tdt and Pdt are the tem-
As a consequence, some assumptions will have to be perature and pressure after the turbine, Put the exhaust
made. They will be justified by a comparison between manifold pressure, Πt , Put is the turbine pressure ratio
P
dt
experimental test data and the results of the model. and γ the specific heat ratio. In this case, the corrected
turbine flow Dt,cor and isentropic efficiency ηt are mapped
2. TURBOCHARGED ENGINE MODEL versus the pressure ratio across the turbine, the corrected
turbocharger shaft speed Nt,cor , and the VGT actuator
position uvgt. As for compressor maps, different methods
Most of the equations governing the behavior of the tur-
have been proposed to obtain these maps from test data
bocompressor can be found in other publications (see for
(see Moraal and Kolmanovsky [1999]). This can be rear-
example Moraal and Kolmanovsky [1999] , Sorenson et al.
ranged in the following form, for more commodity :
[2005] or Eriksson [2006]). The novelty of the approach pre- N P
t dt
sented here lies in the simplification proposed further and the D
t = Dt
φ
(Πt, √ , uvgt)Πt √ (5)
control strategy designed from the simplified model. T T
ut ut
65
2009 IFAC E-CoSM (E-CoSM'09)
Nov 30 - Dec 2, 2009, IFP, Rueil-Malmaison, France.
D
egr,lp LP EGR valve LP EGR Cooler
Variable Geometry
Turbocharger uvgt Exhaust
D
in Oxydation CatalystParticulate Filter Throttle
Dt
P T X D
P , T ut, ut,
out
Puc, Tuc dt dt exh
Exhaust Manifold
Dc
HP EGR
Water−Cooled
Cooler
Air Cooler
HP EGRValve
Degr,hp
Figure 1. Dual mode Diesel-LTC engine architecture including a variable geometry turbocharger, a cooled high-
pressure EGR loop and a cooled low-pressure EGR loop. Sensors are the engine speed, the intake manifold
pressure and temperature (Pdc and Tdc), the compressor upstream pressure and temperature, (Puc and Tuc),
the manifold air flow (Dc).
4
2.3 Intake and exhaust modelling
3
Π [−]
1
temperature, and composition) are assumed to be
homo-geneous. The entire volume between the
0 compressor and the engine can be lumped into a single
0.2
0.15 2.5
volume. The mass balance in this volume and in the
0.1 1.5
2 exhaust manifold leads to
1
0.05 0.5 x 10
5
dPut = RTut (Dasp + Df − Dt − Degr,hp)
0 0 dt Vut (7)
D D D
= ( c + egr,hp − asp)
2.2 Engine modelling dPdc RTdc
dt Vdc
66
2009 IFAC E-CoSM (E-CoSM'09)
Nov 30 - Dec 2, 2009, IFP, Rueil-Malmaison, France.
V
ηv ΠcΨ + Df − D egr, hp ˙ = h(¯z1) (11) 2
z¯
dt
egr,lp c
dc
dPdc RTdc
1−γ γ−1
Thus, the new reference system writes
d 1 2 Nt
,u cP
(9) dt ( 2 JtNt ) = φDt (Πt, √ Tut vgt) p dt Tutηtφt(Πt)
where φt(Π) , Π(1 − Π γ ) and φc(Π) , Π γ − 1. This −(η Π Ψ − D )c T
1
φ
p
(Π )
model takes account of parameters external to the tur- v c egr,hp p uc c c
bocharger itself : temperatures upstream the compressor ηc
and turbine, pressure downstream the turbine. However, it
contains five states. Since the ultimate purpose of this work Nt P dt
η Π Ψ=φ (Π ,√ ,u ) √
is to design a model based control law, further D Dt D t T T
v c ut vgt ut
simplifications have to be undertaken. Different types of +
− f egr,hp
assumptions will be made and verified experimentally.
(12)
3. MODEL REDUCTION The simplified dynamics write now as a first order nonlin-
ear dynamics with an algebraic equation, the steady state
The first type of assumptions concerns the dynamics. solution of the intake and exhaust dynamics
The second type concerns the steady state
dependencies. The purpose is to keep only the relevant 3.2 Steady state assumptions
dynamics of the system, and parameters that can be
measured or estimated from the available sensors.
High pressure flow simplification
3.1 Dynamic simplification : Model simplification by singular Contrary to the tur-
perturbation bocharger, the high pressure EGR loop has a direct con-
trol, i.e. the EGR valve controls directly this flow. Again,
The fifth order nonlinear system (9) accurately describes simplification is made in order to substitute the high
the dynamics of the system. However, one can notice pressure EGR flow by its reference value that is directly
that the turbocharger speed is much slower than the linked to the intake pressure. We introduce variable δHP
pressure dynamics. Indeed, typically we have ≃
V
ut to characterize the choice of EGR circuit : when HP EGR
V
dc
RTut
is used δHP is equal to 1, it is equal to 0 otherwise. With
5e − 9, RTdc ≃ 5e − 8 and Jt = 3e − 5. This suggests this notation, we have :
to simplify these dynamics with a singular perturbation X δ
V
ut D int HP
η
method Khalil [1992]. Let ǫ , RT be a scalar that ut egr,hp = Xexh vol ΠcΨ
represents all the small parameters to be neglected. The where Xint (resp Xexh) is the intake (resp. exhaust)
reference dynamics (9) has the form of the standard
burned gas ratio.
singularly perturbed system
z˙1 = φ(z1, z2, u) Turbine flow simplification The turbine can be consid-
ǫz˙2 = ψ(z1, z2, ǫ) (10) ered as a restriction on the exhaust gas flow. However, the
T standard equation for compressible flow across an orifice
where z1 , Nt, z2 , [ Puc Pdc Put Pdt ] . In other words, cannot be applied in this case. Modified versions of this
we split the slow z1-dynamics (the power balance) and the equation have been proposed which fit better the experi-
z2-fast dynamics (mass balances). mental results, based on various assumptions (see Eriksson
The equation ψ(z1, z2, 0) = 0 has a unique root of interest [2006]). Most of them neglect the influence of the turbine
z2 = h(z1). In details, it is speed. The formula kept in the present case is given below,
P the justification being that it shows a good correlation with
Nt dt the characterization data (see Figure 4).
T T P
ηv ΠcΨ + Df − Degr,hp = φDt (Πt, √ ut , uvgt)Πt √ ut dt
N P
D D φ
out + egr,lp = D (Πt, √ t , uvgt)Πt √ dt Dt = √ φ ψ u
turb(Πt) vgt( vgt)
T
t ut
T T
ut ut
D D D
in + egr,lp = c where ψvgt(uvgt) is equivalent to an effective area (repre-
c + Degr,hp = η v ΠcΨ
D
sented in Figure 3) and
67
2009 IFAC E-CoSM (E-CoSM'09)
Nov 30 - Dec 2, 2009, IFP, Rueil-Malmaison, France.
3
−2 −γ−1 shows experimental measurements. As experimentally rep-
Πt
2
2γ γ γ
φturb(Πt) , √ R
s
γ−1 Πt − Πt
x 1010
4.5
4
1.6
3.5
N [2t rpm 2 ]
1.4 Linear fit
2.5
1.3 2
ψ vgt
1.2 1.5
1.1 1
0.5
1
0
0.9 0.5 1 1.5 2 2.5 3 3.5
Πc [ − ]
0.8
0.7
Figure 5. Experimental results at steady state. Variation
0 10 20 30 40 50 60 70 of the turbocharger speed square Nt2 w.r.t. the
VGT [%]
com-pressor pressure ratio Πc.
0.06
D Tt )t.5 P
68
2009 IFAC E-CoSM (E-CoSM'09)
Nov 30 - Dec 2, 2009, IFP, Rueil-Malmaison, France.
X δ
int HP δ
HP
β = (1 − Xexh )ηv Ψ
α1 = cpTutηt 2 and α2 = cpTuc 1 2
Jta √
ηc Jta
P T D
α3 = √
dt and α4 = P
ut f Πc,sp
M −1
β Tut dt
α1(βΠc,sp + Df )
)
69
2009 IFAC E-CoSM (E-CoSM'09)
Nov 30 - Dec 2, 2009, IFP, Rueil-Malmaison, France.
1.8 70
HP EGR
1.7 60 LP EGR
Manifold Pressure[bar]
1.6
50
1.5
VGT [%]
40
1.4
1.3
Intak
30
1.2
e
20
1.1 HP EGR
LP EGR 10
1
0 20 40 60 80 100 120 140 0
Time [s] 0 20 40 60 80 100 120 140
Time [s]
Figure 7. Intake manifold pressure with HP or LP EGR Figure 9. Turbocharger actuator with HP or LP EGR
6000
0.3
HP EGR ff HP EGR
LP EGR 5000 ff LP EGR
0.25 I HP EGR
4000 I LP EGR
0.2 3000
BGR
2000
0.15
1000
0.1
0
0.05
0 20 40 60 80 100 120 140
−1000
Time [s] 0 20 40 60 80 100 120 140
Time [s]
Figure 8. Estimated intake manifold BGR with HP or LP
EGR Figure 10. Controller terms with HP or LP EGR. I : inte-
gral term represented by C in (15) , ff : feedforward
term equal to α2βψc(Πc,sp) in (15)
REFERENCES . Note that the presented controller terms are in the unit
of the power balance (in Watt).
Eriksson, L. (2006). Modelling and control of turbocharger
si and di engines. In Proceedings of ECOSM conference. of the SAE Conference, 1999-01-0908.
Grondin, O., Moulin, P., and Chauvin, J. (2009). Control Moulin, P., Chauvin, J., and Youssef, B. (2008). Modelling
of a turbocharged engine fitted with high pressure and and control of the air system of a turbocharged gasoline
low pressure exhaust gas recirculation systems. In engine. In Proc. of the IFAC World Conference.
Proc. of the Control and Decision Conf. Shutty, J. (2009). Control strategy optimization for hybrid
Heywood, J. (1988). Internal Combustion Engine Funda- egr engines. In Proc. of the SAE conference, 2009-01-
mentals. McGraw-Hill, Inc. 1451.
Jensen, J., Kristensen, A., Sorensen, S., Houbak, N., Sorenson, S., Hendricks, E., Magnusson, S., and
and Hendricks, E. (1991). Mean value modeling of a Bertelsen, A. (2005). Compact and accurate
small turbocharged Diesel engine. In Proc. of the SAE turbocharger mod-elling for engine control. In Proc. of
Conference, 910070. the SAE Confer-ence.
Khalil, H. (1992). Nonlinear Systems. Prentice-Hall, Inc. Youssef, B., Moulin, P., and Grondin, O. (2007). Model
Kokotovi´c, P., O’Reilly, O., and Khalil, H. (1999). Singular based control of turbochargers : Application to a
Perturbation Methods in Control: Analysis and Design. Diesel HCCI engine. In Proc. of the Conf. on Control
Soc. for Industrial & Applied Math. Applica-tion.
Moraal, P. and Kolmanovsky, I. (1999). Turbocharger
modeling for automotive control applications. In Proc.
70