Preprints of the 18th IFAC World Congress

Milano (Italy) August 28 - September 2, 2011

An MRAC Approach for Tracking and

Ripple Attenuation of the Common Rail
Pressure for GDI Engines
Umberto Montanaro ∗ Alessandro di Gaeta ∗ Veniero Giglio ∗
∗
Istituto Motori, National Research Council, Naples, Italy (e-mails:
{u.montanaro, a.digaeta, v.giglio}@im.cnr.it).

Abstract: Gasoline Direct Injection (GDI) spark ignition engines equipped with the Common
Rail (CR) system strongly improve engine performance in terms of fuel consumption and
pollutant emission reduction. As a drawback the fuel pressure in the rail has to be kept as
constant as possible to the demanded pressure working set-points in order to achieve the
advantages promised by this technology. In this work a Model Reference Adaptive Control
(MRAC) algorithm based on the Minimal Control Synthesis (MCS) strategy is proposed to
reduce the residual pressure in the rail. Numerical results based on a CR mean value model,
previously proposed in the literature and experimentally validated, show that a very satisfactory
attenuation of the pressure ripple as well as pressure tracking are attained in different working
conditions. A quantitative comparison with a classical gain scheduling model-based control
approach confirms furthermore the effectiveness of the proposed adaptive control strategy.

Keywords: Adaptive control, Automotive control, Disturbance rejection, Common Rail, GDI
engine.

1. INTRODUCTION the CR device and driven by the engine camshaft induces

The main objective of a CR system is to feed the electro-
injectors with high fuel pressure independently of the
quantity of fuel to be injected. In fact, high pressure
injections allow a finely atomization of the fuel spray
and promote a better air/fuel mixing, resulting in sig-
nificant combustion improvements HaiFeng et al. (2008);
Tomishima et al. (2008).
Recently the CR injection technology, originally intro-
duced for diesel engines, has been extended to gasoline
engines Corno et al. (2008); Tomforde et al. (2008) in order
to reduce pollutant emissions, as required by the more and
more restrictive emission legislations, and improve engine
performance perceived by customers. More in details, for
spark ignition engines, the CR system is useful to reduce
the exhaust emissions and fuel consumption and to im-
prove the driving dynamics Achleitner et al. (2007).
The precise regulation of the pressure into the rail is the
crucial control task to be solved to take full advantage
of this powerful technology. (For instance, by reducing
the pressure ripple in the manifold, cycle-by-cycle fuel
injection variability is improved and a better control of
the amount of the fuel to inject is attained for the benefit
of pollutant reduction.)
Basically, the pressure ripple (also known in the techni-
cal literature as residual pressure) is caused by pressure
waves due to intermittent openings and closings of the
injectors, which make a constant injection pressure a pure
theoretical assumption. Moreover, the periodic motion of
the High-Pressure (HP) mechanical pump included into
additional periodic pressure oscillations.
Even though different control strategies have been devel-
oped to stabilize the injector pressure in common rail
system (see for example the control strategy based on
the quantitative-feedback theory (QFT) used in Chatla-
tanagulchai et al. (2009) and the hybrid control approach
presented in Balluchi et al. (2006), just to name a few) the
problem of reducing effectively the residual pressure in the
rail is still open.
In order to deal with all the periodic disturbances that
characterize the CR plant, it seems reasonable to imple-
ment a robust adaptive control strategy where the control
gains are properly varied, or better adapted, according to
the actual system behavior. Surprisingly, this approach has
only been marginally investigated in the existing literature
on the control of CR in automotive systems.
The aim of this paper is to confirm numerically the dis-
turbance rejection capability of a MRAC strategy, named
NEMCSI (New Extended Minimal Control Synthesis with
Integral action) algorithm, recently proposed in the liter-
ature di Bernardo et al. (2010). This control strategy is
based on the standard MCS algorithm Stoten and Ben-
choubane (1990) augmented with an integral and robust
action in order to increase tracking performance. As the
standard MCS algorithm, also the NEMCSI strategy relies
on minimal knowledge of the plant and it can be proven
to guarantee asymptotic stability of the error system and
to be robust against rapidly varying disturbances and
unmodelled nonlinear perturbations when the plant is in
control canonical form Stoten and Benchoubane (1990);
di Bernardo et al. (2010).

Copyright by the 4173

International Federation of Automatic Control (IFAC)
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

In this paper we select as CR model the control oriented

model proposed in di Gaeta et al. (2009) for GDI engines
Zhao et al. (2002). Although this model is simpler with
respect to more complex modeling approaches (see for
example Balluchi et al. (2006), Digesu et al. (1994) and
Morselli et al. (2002)), it has been experimentally validated
showing a good prediction of the mean-value rail pressure
and a qualitative behavior of the residual pressure. Notice
that, CR for GDI engines require more sophisticated con-
trol strategies compared to traditional port fuel injection
engines since the injection fuel pressure can vary up to 100
(bar) according to the engine control strategies mapped
into the Electronic Central Unit (ECU).
Fig. 1. Common rail injection system description: 1.
Numerical results for different working conditions show fuel tank; 2. low-pressure pump; 3. fuel filter; 4.
the effectiveness of the proposed adaptive strategy to high-pressure pump; 5. common manifold; 6. electro-
tame the complex dynamics of the CR model, providing injector; 7. pressure sensor; 8. pressure regulation
both excellent pressure tracking performance and residual electro-valve.
ripple attenuation. A comparison with the gain scheduling
model-based control approach proposed in di Gaeta et al.
(2011a) in terms of disturbance reduction confirms the
efficacy of the adaptive method with respect to classical
controllers.
The paper is outlined as follows: in Sec. 2 a description
of the CR plant is briefly presented for the sake of
completeness, in Sec. 3 the NEMCSI control strategy is
described, whereas details on the design of the adaptive
action on the CR system are given in Sec. 4. Numerical Fig. 2. High pressure pump scheme: 1. piston; 2. triangular
results on different cases of interest have been then shown cam; 3. shaft; 4. exit hole.
in Sec. 5 before the conclusions in Sec. 6. In the appendix A
the CR model used for the numerical analysis is reported
to show its nonlinearities and the parameters dependency assured by a triangular cam (three-lobe) connected to the
on the engine speed that make the control objective a not pump’s shaft and each pumping group is characterized by
trivial task. an intake and exhaust valve. The combined action of the
three pumping groups allows to reach a pressure up to
2. COMMON RAIL SYSTEM 100 − 120 bar.
The electro-valve ((8) in Fig. 1) used for the regulation
The CR plant for spark ignition engine, shown in Fig. of the pressure in the rail is driven by electronic circuit
1, is composed mainly by two separated sections: a low- controlled with the PWM strategy.
pressure circuit, consisting of a fuel tank, a fuel feed pump
with a preliminary filter, low-pressure pipes and a fuel The main objective of a CR system is to supply the
filter; an high-pressure circuit formed by an high-pressure electro-injectors with a high pressure fuel independently
pump, an high-pressure line with a pressure sensor, a of the amount of fuel to be injected. The main benefit is,
pressure regulator valve, a flow stopper and an injector. therefore, to decouple the regulation of the high-pressure
The low-pressure electro-pump (2) (p = 3 − 5 bar) forces pump by the functioning of the injectors, differently from
the fuel from the tank (1) toward the high-pressure (HP) traditional injection systems where the mechanical pump
mechanical pump (4) (p = 100 − 120 bar), crossing the generates a pressure that depends on the amount of
filter (3) aimed at cleaning the fuel. The mechanical pump injected fuel.
compresses the fuel and sends it into the common manifold
(5) (named common rail) equipped with the electro-
injectors (6). The manifold is designed in order to filter 3. THE NEMCSI ADAPTIVE ALGORITHM
the oscillations in the fuel due to the high-pressure pump
and the intermittent working of the injectors. Finally, the
pressure in the manifold is regulated through the sensor The NEMCSI algorithm was introduced in di Bernardo
(7) and the electro-valve (8) that flows the excess of fuel et al. (2010) as an extension of the MCS model reference
back into the tank Hongrong et al. (2008). adaptive scheme Stoten and Benchoubane (1990). As the
MCS algorithm, it relies on minimal knowledge of the
The HP pump, shown in Fig. 2, is formed by three plant dynamics. Namely, it is assumed that the plant is
small pistons arranged in radial position (radial-jet) at controllable and it has unknown parameters but a known
an angular distance of 120◦ . The pump is driven by the phase canonical structure, as:
engine through the camshaft and it does not require the
phasing since the instant and time of injection are devoted ẋ(t) = Ax(t) + Bu(t) + d(t, x), (1)
to the ECU, which manages the opening of the injectors. n n
where x(t) ∈ R , u(t) ∈ R, d(t) ∈ R and
The alternating movement of the three small pistons is

4174
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

0

1 0 ... 0
 Z t

 0 0 1 ... 0  xI (t) = [r (τ ) − y (τ )]dτ, y (t) = Cx (t) , (14)

 . 0
.. .. .. . ,
 ..
A= . . . ..  (2) t
Z
KI (t) = α ye (τ )xI (τ ) dτ + βye (t) xI (t) , (15)

 0 0 0 ... 1 
0
−a1 −a2 ... . . . −an Z t

B =(0 0 ··· 0 b) ,
T
(3) KN (t) = γ |ye (τ )| dτ, γ > 0, (16)
0
T
d(t, x) = ( 0 0 · · · 0 f (t, x) ) . (4) being C the output matrix of the plant and ye the output
with f (t, x) being a time and state dependent nonlinear error as in (9). Notice that the signum of the adaptation
function which models the disturbance acting on the weights α and β in (8a), (8c), (15) and γ in (16) has to be
linear part of the plant model and unmodelled nonlinear the same of the term b in (3). Furthermore, typically all the
dynamics. Notice that for the design of the NEMCSI adaptive gains are initialized to zero, i.e. K(0) = K0 = 0,
strategy only the signum of b in (3) has to be known. KR (0) = KR0 = 0, KI (0) = KI0 = 0 and KN (0) = KN0 =
0.
The main aim of NEMCSI control is for the plant states,
x(t), to track asymptotically the states, xm (t), of a given In di Bernardo et al. (2010) it has been proven via
asymptotically stable reference model of the form a passivity based analysis that, as the classical MCS
algorithm, the control action (6) guarantees asymptotic
ẋm (t) = Am xm (t) + Bm r(t), (5) stability of the error dynamics (i.e. xe → 0 when t → ∞)
n
with xm (t) ∈ R , r(t) ∈ R being some desired reference in absence of the d(t, x)-term in (1), and robustness to
signal and Am (Hurwitz matrix) and Bm given in the same unmodelled plant dynamics and disturbance of the form
canonical form as that of the plant. (4) as well as robustness to plant parameter variations.
Moreover the two additional terms, i.e. uN (t) and uI (t)
The control action provided by the NEMCSI algorithm is in (6) have shown to increase tracking and disturbance
given as follows: rejection performance.
u(t) = uM CS (t) + uI (t) + uN (t), (6) We note that, when the reference model in (5) is chosen
where uM CS (t) is the classical MCS input which consists as the a nominal linear model of the plant controlled via
of a feedforward and a feedback action with time-varying a classical LQ (Linear Quadratic) strategy Anderson and
adaptive gains defined as Moore (1971), the control action (6) is termed as LQ-
NEMCSI (see di Bernardo et al. (2010) for further details).
uM CS (t) = K(t)x(t) + KR (t)r(t), (7)
with
4. NEMCSI CONTROL OF THE COMMON RAIL
Z t
K(t) = α ye (τ )xT (τ )dτ + βye (t)xT (t), (8a) The first step for the design of the NEMCSI adaptive
0
controller is to verify if the plant dynamics match or not
K(0) = K0 , K ∈ Rn , (8b) the structure (1). In this work we have chosen from the
Z t
literature the CR model proposed in di Gaeta et al. (2009)
KR (t) = α ye (τ )r(τ )dτ + βye (t)r(t), (8c) that describes the pressure, p(t) (bar), in the CR as a
0
function of the pump speed and electro-valve current, i
KR (0) = KR0 , KR ∈ R (8d) (A), and it can be decomposed into two terms, namely, the
and α and β being scalar adaptation weights. The output mean pressure term, p̄(t) (bar) and the residual pressure,
error ye is computed as η(t) (bar), describing the ripple around the mean value.
The CR dynamical system is then given by
ye (t) = Ce xe (t), (9)  
di R Vb aδ(t) + b
where =− i+ (17a)
dt L L 100
p̄(t) = c(N )i + d(N ), (17b)
xe (t) = xm (t) − x(t), (10a) p(t) = p̄(t) + η(t; p̄, N, Tinj ), (17c)
Ce = [ 0 . . . 0 1 ] P , (10b) where: δ (%) is the duty-cycle expressed in percentage
and P is the solution of the Lyapunov equation terms of the PWM signal used to actuate the electro-valve
((8) in Fig. 1); a and b are parameters of the electro-valve
P Am + ATm P = −M, M > 0. (11) actuation circuit; L (H) and R (Ω) are the electro-valve
inductance and the electric resistance of coil respectively,
The two additional control terms in (6) are given as whereas Vb (V) is the battery voltage supplying the power
follows: circuit; N (rpm) is the rotational speed of the high-
pressure pump that is equal to N = Ne /2 with Ne being
uI (t) = KI (t) xI (t) , (12) the engine speed; Tinj (ms) is the injection time. The terms
c(N ) and d(N ), given in Appendix, are nonlinear functions
uN (t) = KN (t) sgn (ye (t)) , (13) that strongly depend on N . (See Appendix A for further
with details for the model of the residual pressure η used in the
numerical validation.)

4175
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

After simple algebraic manipulations the dynamical pres-

sure model
" is #
∂c Ṅ R aVb
ṗ = − p+c u + Ψ (t; p̄, N, Tinj ) , (18)
∂N c L 100L
where u = δ is the control input and Ψ is the nonlinear
function given in Appendix A.
It is apparent by (18) that the CR model satisfies the
matching condition required by the NEMCSI algorithm
hence this adaptive control strategy is applicable.
As reference model, here we select for the NEMCSI imple-
mentation a first order LTI model with a rise time of about
100 (ms) in order to satisfy the requirements on the CR Fig. 3. Graphical representation of the Attenuation factor.
response given in di Gaeta et al. (2011a). In implementing
the control action (6), the following issues have been con- 100
sidered: (i) as it usually happens when implementing the 90
MCS strategy Stoten and Benchoubane (1990), the adap- 80
tive weights α, β and γ, which modulate the adaptive gains
70
in (8a), (8c), (15) and (16), have been chosen heuristically

P, Pr [bar]
60
as a tradeoff between convergence time and reactivity of
the control action; (ii) locking of the adaptive gains was 50

implemented, in particular, the adaptive gain KN is frozen 40

when it exceeds 10 (%) of duty-cycle; (iii) the presence of 30

the switching function in (13) introduces chattering (high 20
frequency switching of the control signal), and in order to 10
avoid this unwanted phenomena, the discontinuous control 0 5 10 15
t [s]
20 25 30

action has been smoothed as in di Bernardo et al. (2010). (a)

Finally, all the integrators of the NEMCSI control action
(6) have been discretized according to the Tustin method 100
with a sampling time of 1 (ms) that is implementable in 90
commercial ECUs. 80

We remark that no a priori knowledge in terms of plant 70

P, P [bar]

parameters or plant nonlinearities has used to design the 60

controller.
r

50

40
5. NUMERICAL RESULTS
30

The NEMCSI adaptive algorithm designed in Sec. 4 has
been widely tested on the CR model (A.1) (or equivalently
(A.4a)-(A.4b)) in order to prove its effectiveness in control-
ling the CR dynamics in terms of regulation, tracking and
disturbance attenuation. In particular, assuming that a
periodic solution of period T̄ is achieved after the time in-
stant t⋆ when the input reference to the closed loop system
is a constant pressure Pr under the periodic disturbance
excitation η(t) (see equation (A.3) in Appendix A), we
measure quantitatively the degree of the attenuation of the
disturbance η according to the attenuation factor defined
as follows:
∆η − ∆p
Af = 100 , (19)
∆η
where ∆p = [maxΦ (p(t) − Pr ) − minΦ (p(t) − Pr )] is the
maximum variation of the pressure in closed loop around
the pressure working point Pr and computed over a
generic time interval Φ , [t1 ; t2 ] so that t1 > t⋆ and
t2 − t1 > T̄ . Analogously, ∆η = [maxΦ (η) − minΦ (η)] is
maximum variation of the pressure disturbance due to
the HP pump and fuel injections computed on the same
time interval. Obviously, higher Af values are obtained for
effective disturbance rejections, instead negative values of
Af denote that the residual/ripple pressure is amplified in
20
10
70 80 90 100 110 120
t [s]
(b)
Fig. 4. NEMCSI Control results for the long manoeuvre.
Rail pressure (blue line), reference pressure (red line)
on the time interval: (a) [0; 30] (s) and (b) [62; 125] (s).
closed loop (see also Fig. 3 for a graphical representation
of ∆p , ∆η and Φ).
In what follows some remarkable results are briefly de-
scribed.
• As first case we consider a long manoeuvre composed by
a sequence of pressure steps. In this case the engine speed is
fixed at Ne = 1000 (rpm) while the injection time for each
fixed reference pressure is listed in Tab. 1. The excellent
tracking is shown in Figs. 4a and 4b. (Notice that from 30
(s) to 62 (s) the reference pressure is kept constant at 30
(bar).)
For each pressure working set-point the attenuation factor
Af has been computed. Our analysis has shown that for
each constant set point Pr a satisfactory attenuation factor

4176
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

Table 1. 10
Pr = 25
Simulation Pr = 35
interval (s) 0-52 52-62.5 62.5-80 80-95 95-112.5 112.5-125 9
Pr = 45
Tinj (ms) 0 8 4 2 4 2
8 Pr = 55
Pr = 65
6 43.5 7 Pr = 75

Tinj [ms]
4 43 Pr = 85
42.5 6
2
P−15[bar], d [bar]

42
0 5
δ [%]

41.5

−2 41
4
−4 40.5

40
−6 3
39.5
1000 2000 3000 4000 5000 6000
2.65 2.7 2.75 2.8
t [s]
2.85 2.9 2.95 2.65 2.7 2.75 2.8
t [s]
2.85 2.9 2.95 Ne [rpm]

(a) (b)
Fig. 6. Injection time for each working point belonging to
8
17
P × Ne .
6

4 16
85
2
P−100[bar], d [bar]

15
0
δ [%]

80
−2 14

−4
13
−6 75
−8 12

−10
11
Af [%] 70
118.65 118.7 118.75 118.8 118.85 118.9 118.95 119 118.65 118.7 118.75 118.8 118.85 118.9 118.95 119 Pr = 25
t [s] t [s]
65 Pr = 35
(c) (d) Pr = 45
60 Pr = 55
Fig. 5. NEMCSI Control results for the long manoeuvre Pr = 65
for some steady state. Residual fuel pressure p(t) − Pr 55 Pr = 75
(blue solid line) and disturbance η acting on the plant Pr = 85
(red dashed line) when the reference pressure is: (a) 50
1000 2000 3000 4000 5000 6000
Pr = 15 (bar) and Tinj = 0 (ms); (c) Pr = 100 (bar) Ne [rpm]
and Tinj = 2 (ms); (b) control action for the case (a);
(d) control action for the case (c). Fig. 7. NEMCSI control results. Attenuation factor for
each working point belonging to P × Ne .
is attained, i.e. Af ∈ [80.92; 92.46] for the entire manoeu-
vre. To confirm the disturbance rejection performance, in out in Bursi et al. (2010), discretization implies the need
Fig. 5 detailed time histories of the residual closed loop of some bounds on the magnitude of the adaptive weights
pressure, the control action and the rejected disturbance (α and β) to avoid unwanted closed loop dynamics then,
η are shown in the case Pr = 15 (bar) (minimum refer- as consequence, the control performance deteriorates. Nev-
ence pressure over the manoeuvre) and Pr = 100 (bar) ertheless, the results here obtained with a sampling time
(maximum reference pressure over the manoeuvre). implementable in the experimental setup used in di Gaeta
• To prove the effectiveness of the NEMCSCI algorithm in et al. (2011a) are very satisfactory if we consider that not
attenuating the pressure disturbance in different working any a priori knowledge of plant dynamics have been used
conditions of reference pressure and engine speed, the for the control design. Results related to the continuous
attenuation factor (19) has been computed for each pair implementation of the control strategy are not reported
(Pr , Ne ) ∈ P × Ne with P , {25 + 10k, k = 0 . . . 6} and here for the sake of brevity.
Ne , {1000 + 500j, j = 0 . . . 10}. For each working point Details on the time history of the residual closed loop
the injection time is that to guarantee the stoichiometric pressure are given for the sake of completeness together
regime of the 2-liters GDI engine, presented in di Gaeta with the control action and the rejected disturbance η
et al. (2010, 2011b), at full load conditions (Wide Open in Fig. 8 in the case (Pr , Ne ) = (35 (bar), 2000(rpm))
Throttle, WOT). Fig. 6 and Fig. 7 show the injection (low engine speed and reference pressure) and (Pr , Ne ) =
time and the attenuation factor for each pair belonging to (65 (bar), 6000 (rpm)) (high engine speed and reference
P × Ne respectively. As it was expected, Fig. 7 shows that pressure). In the latter case the control action in Fig.
the attenuation factor is a decreasing function of the CR 8d appears less smooth than the previous cases since
pressure and engine speed. Nevertheless, the performance the sampling frequency (1 (kHz)) is comparable with the
index (19) never goes down to 55% and it remains greater highest disturbance frequency (200 (Hz)) due to injectors
than 70% independently of the pressure as far as the engine (see Appendix A) at this engine speed. Nevertheless the
speed is below 2500 (rpm). adaptive strategy still works effectively.
We remark that much better attenuation factors can be In order to stress the efficiency of the NEMCSI control
achieved if the control action (6) is not discretized, or dis- action, the controller proposed in di Gaeta et al. (2011a)
cretized with a smaller sampling time. In fact, as pointed has been tested on the same working points. That control

4177
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

12
39
8
2000
6 38 10

4
37
8 1500
2
P−35[bar], d [bar]

Tinj [ms]
36

N [rpm]
0
6

δ [%]

e
−2 35 1000
−4 4
34
−6
500
−8 33 2

−10 32
0 0
−12 0 1 2 3 4 5 0 1 2 3 4 5
31 t [s] t [s]
18.32 18.34 18.36 18.38 18.4 18.42 18.44 18.46 18.48 18.32 18.34 18.36 18.38 18.4 18.42 18.44 18.46 18.48
t [s] t [s]
(a) (b)
(a) (b)
Fig. 10. Experimental traces of (a) injection time and (b)
8
36

34
engine speed measured in a commercial ECU of a 2-
6
32
liters GDI engine.
4

2 30
P−65[bar], d [bar]

0 28
δ [%]

−2
26
−4
24 120
−6
22
−8

−10 20
100
−12 18
18.44 18.45 18.46 18.47 18.48 18.49 18.5 18.44 18.45 18.46 18.47 18.48 18.49 18.5
t [s] t [s]

80
(c) (d)

Fig. 8. NEMCSI control results in steady state. Residual P, P [bar]

r
60
fuel pressure p(t)−Pr (blue solid line) and disturbance
η acting on the plant (red dashed line) when: (a) the 40
reference pressure is Pr = 35 (bar) and Ne = 2000
(rpm); (c) the reference pressure is Pr = 65 (bar) and 20
Ne = 6000 (rpm); (b) control action for the case (a);
(d) control action for the case (c). 0 1 2 3 4 5
t [s]

25 Fig. 11. NEMCSI control results. Tracking of an experi-

P = 25
20
r
Pr = 35
mental reference pressure at the engine startup: refer-
P = 45 ence pressure (red line) and controlled pressure (blue
r
15
P = 55
r
line).
P = 65
10 r
P = 75
r
A [%]

5 P = 85
r
• As last case we test here the pressure tracking perfor-
f

0 mance of the proposed adaptive control scheme at the

−5
−10
−15
1000 2000
Fig. 9. Gain scheduling control results. Attenuation factor
for each working point belonging to P × Ne .
action is basically a model-based feed-forward compen-
sator coupled with an integral controller whose gains are
scheduled on the basis of the engine speed. In particular for
each engine speed, model (A.1) has been used to provide
in closed loop a certain nominal behavior independently
on the working engine speed. Fig. 9 shows that poor
disturbance rejection performance are obtained when this
classical approach is used. In particular we note that for
engine speeds greater than 1500 (rpm) the attenuation
factor takes negative values independently of the reference
pressure. Hence, for a wide range of working conditions
the control action even amplified the disturbance acting
on the plant.
engine startup (cranking). In particular, the reference CR
pressure, the injection time as well as the engine speed are
those experimentally obtained with a 2-liters GDI engine
during a cold start. The experimental injection time Tinj
3000 4000 5000 6000 imposed by the ECU and the experimental engine speed
Ne [rpm]
Ne are reported in Figs. 10a and 10b, respectively.
Fig. 11 shows a very satisfactory tracking of the ex-
perimental reference pressure provided by the NEMCSI
strategy. In particular tracking is not lost over the time
interval [0.8, 1.5] (s) despite of an abrupt increasing of
the reference pressure, engine speed as well as injection
time. More in detail, in this small time range the reference
pressure sweeps the entire pressure range [10, 110] (bar)
simultaneously the engine speed varies up to 2200 (rpm)
and the injection time reaches the high value of 12 (ms).
Figs. 12a and 12b show, for the sake of completeness, the
control action and adaptive gains. In particular all the
gains remain bounded also for this hard test manoeuvre.
We note the control action does not exhibit chattering as
it might appear by Fig. 12a. This can be confirmed by a
zoom on the time interval [1.5; 1.8] (s) not reported here
for the sake of brevity.

4178
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

100 Anderson, B.D.O. and Moore, J.B. (1971). Linear Optimal

90 Control. Prentice Hall, Englewood Cliff, N.J., USA.
ISBN: 9780201808681.
80
Balluchi, A., Bicchi, A., Mazzi, E., Vincentelli, A.L.S.,
70 and Serra, G. (2006). Hybrid modelling and control
60 of the common rail injection system. In J. Hespanha
and A. Tiwari (eds.), Hybrid Systems: Computation and
δ [%]

50
Control, volume 3927 of Lecture Notes in Computer
40
Science, 79–92. Springer Berlin / Heidelberg.
30 Bursi, O., Stoten, D.P., Tondini, N., and Vulcan, L. (2010).
20 Stability and accuracy analysis of a discrete model
10
reference adaptive controller without and with time
delay. International Journal for Numerical Methods in
0
0 1 2 3 4 5 Engineering, 82(9), 1158–1179.
t [s]
Chatlatanagulchai, W., Aroonsrisopon, T., and Wanna-
(a) ton, K. (2009). Robust common-rail pressure control
for a diesel-dual-fuel engine using QFT-based controller.
1.5
SAE Technical Paper, (2009-01-1799).
Corno, M., Savaresi, S.M., Scattolini, R., Comignaghi, E.,
1 Sofia, M., Palma, A., and Sepe, E. (2008). Modelling,
parameter identification and dynamics analysis of a
0.5 common rail injection system for gasoline engines. In
adaptive gains

Proc. of the 17th IFAC World Congress.

0 di Bernardo, M., di Gaeta, A., Montanaro, U., and Santini,
−0.5
−1
−1.5
0 1 2 3 4 5
t [s]
(b)
Fig. 12. NEMCSI control results. Tracking of an experi-
mental reference pressure at the engine startup: (a)
control action, (b) adaptive gains, K (blue line), Kr
(red line), KI (magenta line) and 10 · KN (back line).
6. CONCLUSION
In this paper we have investigated an adaptive technique
for the control of the fuel pressure into a common rail for
GDI engine application. The adaptive strategy is based
on the so called minimal control synthesis algorithm and
require a minimal knowledge of the plant dynamics to be
controlled. Numerical results have shown that the adaptive
controller is able to cope with plant parameter variations
caused by engine speed variation and nonlinear periodic
pressure state dependent disturbances. This is particular
important for the common rail system to attain better
engine performance.
The promising numerical results here obtained both in
terms of tracking and disturbance attenuation for differ-
ent steady and transient experimental working conditions
(e.g. engine startup) push us in the next future to test
experimentally the adaptive control strategy designed in
the paper.
REFERENCES
Achleitner, E., Bcker, H., and Funaioli, A. (2007). Direct
injection systems for otto engines. SAE Tecnical Paper,
(2007-01-1416).
S. (2010). Synthesis and experimental validation of the
novel LQ-NEMCSI adaptive strategy on an electronic
throttle valve. IEEE Transactions on Control Systems
Technology, 8(6), 1325–1337.
di Gaeta, A., Fiengo, G., Palladino, A., and Giglio, V.
(2009). A control oriented model of a common-rail
system for gasoline direct injection engine. In Proc. of
48th IEEE Conference on Decision and Control & 28th
Chinese Control Conference.
di Gaeta, A., Fiengo, G., Palladino, A., and Giglio, V.
(2011a). Design and experimental validation of a model-
based injection pressure controller in a common rail
system for gdi engine. In Proc. of American Control
Conference.
di Gaeta, A., Montanaro, U., and Giglio, V. (2010). Idle
speed control of GDI-SI engines via ECU-1D engine co-
simulation. SAE Technical Paper, (2010-01-2220).
di Gaeta, A., Montanaro, U., and Giglio, V. (2011b).
Model-based control of the afr for gdi engines via ad-
vanced co-simulation: an approach to reduce the devel-
opment cycle of engine control systems. ASME Journal
of Dynamic Systems, Measurement, and Control. Ac-
cepted for pubblication.
Digesu, P., Ficarella, A., Laforgia, D., Bruni, G., and
Ricco, M. (1994). Diesel electro-injector: A numerical
simulation code. SAE Technical Paper, (940193).
HaiFeng, S., YouTong, Z., Jun, W., and LianDa, L. (2008).
Researches of common-rail diesel engine emission con-
trol based on cylinder pressure feedback. In Proc. of
IEEE Vehicle Power and Propulsion Conference.
Hongrong, W., Youtong, Z., and Jun, W. (2008). Studies
of control strategies for high pressure common rail diesel
engine. In Proc. of IEEE Vehicle Power and Propulsion
Conference.
Morselli, R., Corti, E., and Rizzoni, G. (2002). Energy
based model of a common rail injector. In Proc. of
International Conference on Control Applications.
Stoten, D.P. and Benchoubane, H. (1990). Robustness of
a minimal controller synthesis algorithm. International

4179
Preprints of the 18th IFAC World Congress
Milano (Italy) August 28 - September 2, 2011

(a) (b)

Fig. A.1. CR plant parameters which are function of the

HP pump speed: (a) c(N ) and (b) d(N ).
Journal of Control, 51(4), 851–861. Fig. A.2. Block diagram of model (A.1).
Tomforde, M., Jeinsch, T., Blath, J., and Dnow, H. (2008). to the motion of the HP pump, say ηP , and another term,
Modelling of a fuel supply system for model-based say ηI , due to the fuel injection. It is easy to prove that
calibration. In Proc. of the 17th IFAC World Congress. both terms are periodic since: ηP (t) is originated by the
Tomishima, H., Matsumoto, T., Oki, M., and Nagata, K. motion of the high-pressure pump, performing a revolution
(2008). The advanced diesel common rail system for every two rounds of the engine shaft; ηI (t) related to the
achieving a good balance between ecology and economy. injectors functioning, activated sequentially one time for
SAE Tecnical Paper, (2008-28-0017). each engine cycle. Hence, the fundamental harmonics are
Zhao, F., Harrington, D.L., and Lai, M.C.D. (2002). Au- ωP = N/30 [rad/s] and ωI = nN/30 [rad/s] for ηP (t)
tomotive Gasoline Direct-Injection Engines. SAE Inter- and ηI (t), respectively, with n being the number of the
national. ISBN: 978-0-7680-0882-1. injectors.
In this paper we model the pressure disturbance η(t) ac-
Appendix A. COMMON RAIL MATHEMATICAL cording to the experimental frequency analysis presented
MODEL in di Gaeta et al. (2011a) but considering only the funda-
mental harmonics of ηP (t) and ηI (t) signals. Hence, the
The model proposed and experimental validated in di Gaeta pressure ripple is described as
et al. (2009) describes the electric dynamics of the electro-
valve, neglecting the effects due to the movement of its η(t; N, Tinj , p̄) = ηP (t; N, p̄) + ηI (t; N, Tinj , p̄), (A.3)
plunger (i.e. inductance variations and back-electromotive with
force), and the actuation circuit used to drive the electro- h  
πN
i  
πN
valve as well. The CR dynamic system is then given by ηP (t; N, p̄) = p̄e−p̄/ν M1 + M2 sgn sin t sin t ,
30 30
πN
h  i
ηI (t; N, Tinj , p̄) = −ξ (Tinj ) + (AI p̄) sin n t + φI H (Tinj ) ,
 
di R Vb aδ(t) + b 30
=− i+ (A.1a)
dt L L 100 where ξ(Tinj ) is a proper function (so that ξ(Tinj ) ≥ 0 with
p̄(t) = c(N )i + d(N ), (A.1b) dξ(Tinj )
> 0 ∀Tinj ) describing the pressure decreases due
dTinj
p(t) = p̄(t) + η(t), (A.1c)
to the injectors, sgn(·) is the signum function, H(·) is the
where: δ (%) is the PWM duty-cycle expressed in per- Heaviside step function, and the constant parameters ν,
centage terms, i (A) is the current, L (H) and R (Ω) are M1 , M2 , AI , φI are tuned to mimic experimental pressure
inductance and the electric resistance of coil respectively, oscillations.
a and b are parameters of the actuation circuit and Vb
(V) is the battery voltage supplying the power circuit; From the previous discussion, model (A.1) can be then
p (bar) is the instantaneous rail pressure, η (bar) is the represented as in Fig. A.2 that highlights η (both its
pressure ripple around the mean value p̄ (bar), and N is the components) as a disturbance acting on the plant output
rotational speed of the high-pressure pump that is equal to be suppressed.
to N = Ne /2 with Ne being the engine speed. By means of simple algebraic manipulations the model
The functions c(·) and d(·) in (A.1) depend on the HP (A.1) can be rewritten in control canonical form required
pump speed N and are modeled as third order polynomials by the NEMCSI algorithm as
" #
of the form: ∂c Ṅ R aVb
3  k 3  k ṗ = − p + c(N ) δ + Ψ (t) (A.4a)
X N X N ∂N c(N ) L 100L
c(N ) = ck , d(N ) = dk . (A.2)
103 103 " #
k=0 k=0 ∂c Ṅ R
The graphical representation in Fig. A.1 shows that these Ψ(t) = − − (η(t) + d(N )) +
∂N c(N ) L
functions strongly vary with N , and then with the engine
speed. ∂d c(N )bVb
+ Ṅ + + η̇(t). (A.4b)
∂N 100L
The η-term in (A.1c) takes into account the wave motion of
the fuel into the rail causing variation of the pressure and it
can be supposed to be composed by two terms, a term due

4180