Civil Engineering Reference
In-Depth Information
is displaced towards the future and the controller with new measurements solves and
updates the optimal control problem.
Additionally in this chapter, differentMPCapproaches have been developed trying
to find a tradeoff between performance and cost and using as manipulated variables
only the fancoil fan power or the fancoil fan power plus the water flow valve.
Some of them use linear controllers based on LTI models of the HVAC system
to directly generate the control signal (fancoil fan power) or to produce temperature
setpoints within a hierarchical control strategy aimed at reaching the highest thermal
comfort level. Different cost functions have been evaluated.
A nonlinear control architecture that makes use of a nonlinear model based on first
principles has been also developed and implemented within a hierarchical control
approach. In this case, it has been considered that the HVAC system has two degrees
of freedom, the fancoil fan power and the water flow through it. The higher layer of
the hierarchical controller contains an optimiser which provides an optimal impulse
air temperature. The lower layer receives as input the impulse air temperature setpoint
and includes a control algorithm which tries to efficiently lead the fancoil unit to this
value.
Real tests have demonstrated the feasibility and performance of these approaches,
both for summer andwinter periods. The chapter has also introduced a novel approach
to perform a multivariable nonlinear control for both thermal comfort and indoor air
quality, using a nonlinear MPC approach. The approach is able to maintain both
thermal comfort and indoor air quality and, at the same time, to optimise the use
of forced and natural ventilation. The use of multiobjective control approaches has
been also discussed.
The chapter extends the previous results dealing with control of a room to the
centralised control of a building (considering it as a set of rooms). In this case, the
challenge arises when the energy demanded by the occupied rooms in the building
is greater than the one available, which is a situation that may occur when using
renewable energy sources or the daily amount of energy is limited.
All the approaches have been commented in terms of benefits and drawbacks and
they can be used depending on the installed network of sensors and actuators and
control objectives.
References
Ahmed O, Mitchel J, Klein S (1996) Application of general regression neural networks (GRNN)
in HVAC process identification and control. ASHRAE Trans 102:1147-1156
Akyildiz IF, Su W, Cayirci E, Sankarasubramaniam Y (2002) Wireless sensor networks: a survey.
Comput Netw 38:393-422
Álvarez JD, Redondo JL, Camponogara E, Normey-Rico J, Berenguel M, Ortigosa PM (2013)
Optimizing building comfort temperature regulation via model predictive control. Energy Build
57:361-372
Åström KJ, Hägglund T (2005) Advanced PID control, ISA, The Instrumentation, Systems, and
Automation Society, Research Triangle Park, NC, p 27709
