Civil Engineering Reference
In-Depth Information
Table 5.3
Summary of the results obtained from each strategy
Strategy
Index 1
a
Index 2
b
Index 3
c
Lambda (
ˉ
)
(A)
0.1
30
30.13
40
0.3
62
18.20
24
(B)
0.0065
58
42.50
56
0.008
84
37.74
50
(C)
0.001
42
42.14
55
0.00355
35
29.39
39
(D)
0.012
47
33.26
44
0.2
57
41.61
55
a
Mean number of changes per hour [-]
b
Percentage of total time in which HVAC system is connected [%]
c
Mean energy consumption per hour [Wh]
measures and a complete model of the fancoil. For this reason, in the following
section a nonlinear control architecture that makes use of a nonlinear model based
on first principles is developed. Another conclusion of these preliminary tests is that
the sample time used for control purposes should be decreased. The control action
experiments stepwise changes whose amplitude can be smoothed by reducing the
sample time. Thus, in the following control approaches, the sample time is decreased.
5.4 An Advanced Control System: A Nonlinear Controller
for Users' Thermal Comfort
Several control strategies for indoor comfort, which can be found in the literature,
are limited to on/off and conventional PID methods (Calvino et al.
2010
) with the
classical limitations that arise with these kinds of controllers, i.e. they are not suitable
in both, counteracting the disturbances, and taking into account energy saving at time
to reach indoor thermal comfort. On the other hand, advanced control strategies have
been proposed in the literature with the aim of overcoming the limitations of these
classical controllers. The most proposed technique to control indoor air temperature
and thermal comfort is MPC (Ferreira et al.
2012
; Freire et al.
2006
,
2008
;Kelman
et al.
2013
;Maetal.
2011
,
2012b
; Ma and Borrelli
2012
; Oldewurtel et al.
2010a
,
b
; Zhang et al.
2013
). However, although the indoor air temperature process has
nonlinear dynamics, most of MPC applications are based on the use of linear models
except for a few works (Ferreira et al.
2012
).
In order to deal with these problems, in this section a control approach based on a
Nonlinear Model-based Predictive Control (NMPC) is explained. Its general archi-
tecture can be observed in Fig.
5.20
. Moreover, the main differences of this approach
with the ones presented in Sect.
5.3
lie in the use of a nonlinear first principles model
which accurately reflects the dynamics of the room climate and takes into account
the main disturbances. Besides, in this case, it has been considered that the HVAC
system has two degrees of freedom, the fancoil power and the water flow through
