Implementation of a predictive simulation-based controller for environmental systems in buildings / Matthias Schuß
VerfasserSchuß, Matthias
Begutachter / BegutachterinMahdavi, Ardeshir ; Kabele, Karel
UmfangVII, 110 Bl. : Ill., graph. Darst.
HochschulschriftWien, Techn. Univ., Diss., 2011
Zsfassung in dt. Sprache
Bibl. ReferenzOeBB
Schlagwörter (DE)Prediktive Regelung / Gebäudesimulation / Langzeittest
Schlagwörter (EN)predictive control / building simulation / long-term study
Schlagwörter (GND)Haustechnik / Energieverbrauch / Simulation / Prädiktive Regelung
URNurn:nbn:at:at-ubtuw:1-40378 Persistent Identifier (URN)
 Das Werk ist frei verfügbar
Implementation of a predictive simulation-based controller for environmental systems in buildings [38.59 mb]
Zusammenfassung (Englisch)

Buildings play an important role in the worlds total energy use. A major part of the energy use of buildings is caused by operation of related environmental systems to provide suitable indoor environmental conditions, including thermal comfort. With increased complexity and mutual interferences of these environmental systems, strategies for optimized control are required. The concept of a predictive simulation-based controller focuses on the holistic optimized control of all these components. Core feature of this controller is an embedded multi-domain simulation kernel to predict the performance of control alternatives, which are generated by a procedure that deploys genetic algorithms. Two test facilities were selected to implement this control concept in realistic office setups. Three rooms in a more then hundred-year-old university building in Vienna/Austria and two rooms in a new office building in Styria/Austria were adapted for this purpose. The controller concept was integrated into the test environments with conventional off the shelf building automation hardware. Control software was programmed to operate the test facilities continuously. Additional, a web based graphical user interface was developed to inform users about the controller state, enable manual control override, and visualize monitored data.

External and internal climate data was collected at both facilities toward an objective assessment and evaluation of the system's performance during test operation in two summer periods (2009 and 2010).

The results of these tests showed the potential of the proposed predictive simulation-based control. Overall, the implementation demonstrated that the proposed control approach can be realized in existing buildings and be integrated into legacy automation systems.