We define building systems as systems that not only consist of the load bearing building structure and its environment, the service systems (installations), and the control systems, but also include building usage and users. Our ultimate goal is a virtual building laboratory for the concurrent simulation of all components of such building systems. This will be an extension of performance simulators. In this paper we focus on an efficient modeling process for deriving object-oriented structural models for virtual building laboratories, with the specialization towards the development, test, and maintenance of integrated control systems. We also focus on control systems that support individual user environments and that can easily be adapted to changes in the use of building systems. For the purpose of efficient modeling, we divide the building system domain into five domains, called building, service, control, functional unit, and user role domain. We present corresponding domain models that predefine domain dependent modeling primitives for project specific building system models. All models are based on the notion of space and matter. In each domain, spaces have special semantics and build aggregation hierarchies, which we call “backbones”. The backbones of the domain models are linked by requirement and spatial relations, thus composing one integrated building system model. INTRODUCTION Nowadays, building simulation is mainly used during the planning process for improving energy efficiency. For this purpose, specific models have been developed for computing heat flow, gains, and losses. This kind of simulation is called performance simulation. More advanced models allow the simulation of long wavelength radiation, light, sound, air flow and service systems for heating and air conditioning. Simulators based on these advanced models can also be used for planning and improving the internal environment for work places or living spaces and for testing building control systems. Performance simulators do not include the inhabitants of buildings, more precisely the organization of the users and the user activities. New modeling approaches have included these entities, but only for the purpose of requirements descriptions and not for simulation. Also, only very simple building control algorithms have been included in performance simulators. Control systems in contemporary buildings are more complex than what is available in performance simulators. There is a demand for even more complex integrated building automation systems. We have shown how such control systems can be efficiently developed (Queins et al., 2002). One of the difficulties that has to be solved by control systems is the increasing and changing demand of the users to take part in the control of their individual personal environment. Another difficulty is the fact that the use of commercial buildings frequently changes and control systems have to be flexible enough or easy to adjust to allow for such changes. During the development and maintenance of building control systems that support individually controllable environments and change of building use, we need an environment for validation (“Do we build the right system?”) and verification (“Do we build the system right?”). This requires more than performance simulation. It requires a view of a buildings as a system, including the building itself (the load bearing structure), the service systems (installations), the functional units (work places, activity spaces), and the inhabitants or users themselves (organization). To complete the building as a system, in short building system, we include the control system. Because of the differences of these five components of a building system, we define five corresponding domains. The result of our current work, which we present in this paper, is a modeling approach that supports the process of deriving models in each of the five domains and of connecting the five models by relations. Models are derived on a project by project basis and thus the modeling process has to be very efficient. Since the goal is an environment for validating and verifying control systems by prototyping, all models have to be transformed into computer programs. Therefore, we use model notations that can automatically be transformed into executable programs. With additional features that support experimentation, the envisioned simulation environment will become a virtual building laboratory (Mahdavi et al., 2002). Besides the testing of control systems, such a virtual laboratory can also be used for virtual construction MODELING THE BUILDING AS A SYSTEM Gerhard Zimmermann University of Kaiserslautern 67653 Kaiserslautern Eighth International IBPSA Conference Eindhoven, Netherlands August 11-14, 2003
[1]
Andreas Metzger,et al.
Towards a Virtual Laboratory for Building Performance and Control
,
2002
.
[2]
Alexander Koutamanis,et al.
Route analysis in complex buildings
,
2001
.
[3]
Robert Lake,et al.
Near-Real-Time Control of Human Figure Models
,
1987,
IEEE Computer Graphics and Applications.
[4]
Anders Olsen,et al.
Systems Engineering Using SDL-92
,
1994
.
[5]
Andreas Metzger,et al.
Specifying Building Automation Systems wigh PROBanD, a Method Based on Prototyping, Reuse, and Object-orientation
,
2001,
OMER.
[6]
A. Hendricx.
Shape, Space And Building Element: Development of a Conceptual Object Model for the Design Process
,
1997,
Proceedings of the 15th International Conference on Education and Research in Computer Aided Architectural Design in Europe (eCAADe).
[7]
Jan Peter Riegel,et al.
Pattern-Based Generation of Customized, Flexible Building Simulators
,
1997
.
[8]
Konstantinos Papamichael,et al.
Building design advisor: automated integration of multiple simulation tools
,
1997
.
[9]
Andreas Metzger,et al.
Model-Based Generation of SDL Specifications for the Early Prototyping of Reactive Systems
,
2002,
SAM.
[10]
Robert Woodbury,et al.
Software Environment to Support Early Phases in Building Design (SEED): Overview
,
1995
.
[11]
Gerhard Zimmermann.
Efficient creation of building performance simulators using automatic code generation
,
2002
.