MULTI-AGENTS FOR ENERGY EFFICIENT COMFORT
Agents for the Energy Infrastructure of the Built Environment: Flexergy
Wim Zeiler, Rinus van Houten, Gert Boxem
Technische Universiteit Eindhoven (TU/e), Vertigo 6.28, P.O. Box 513, 5600 MB Eindhoven,The Netherlands
Joep van der Velden, Willem Wortel, Jan-Fokko Haan, Paul Noom
Kropman Building Services, Nijmegen, The Netherlands
Rene Kamhuis, Maarten Hommelberg
Energy Research Centre Netherlands (ECN), Petten, The Netherlands
Henk Broekhuizen
Installect, Nijkerk, The Netherlands
Keywords: Multi-agents, Built environment, Building Services, Flexergy.
Abstract: Synergy between end-user, building and the built environment is the ultimate in the intelligent comfort
process control concept. This new comfort control technology is based on the use of agent technology and
can further reduce energy consumption of buildings while at the same time improve individual comfort. The
TU/e (Technische Universiteit Eindhoven) together with Kropman and ECN (Energy research Centre
Netherlands) work together in the research for user based preference indoor climate control technology.
Central in this approach is the whole building design process including the energy infrastructure which
makes it possible to reduce energy consumption by tuning demand and supply of the energy needed to fulfil
the comfort demand of the occupants of not just one building but a set of physical or virtual connected
buildings.
1 INTRODUCTION
There is a persistent discrepancy between increasing
demands for comfort in buildings and the need to
decrease use of energy. In Europe comfort in
buildings needs 40% of the total energy. With
effects of Global warming becoming more and more
apparent (Alley et al., 2007) there is a need to reduce
this energy demand by comfort within the built
environment. Over the years energy efficiency of
buildings has increased. At first by better ways of
constructing followed by applying better insulation
and better glazing. Also the introduction of more
efficient building equipment has lead to further
reduction of the energy use of buildings.
Present control systems for office buildings
already make use of new technical possibilities
offered by computer networks from the Building
Management System (BMS) about the users, e.g.
comfort demands or comfort preferences of the
building occupants.
New comfort control technology, such as
individual control, offers new possibilities to further
reduce energy consumption of office buildings.
Dynamic online steering of individual comfort
management and building management could save
up to 20% of current energy consumption
(Akkermans, 2002). The behaviour of building
occupants needs to be taken into account as it is
responsible for almost half the outcome of planned
energy reduction (Claeson-Jonsson, 2005).
As until now the individual comfort profile of
each user has not been part of the building comfort
system control strategy in offices. New
technological development is needed to incorporate
the behavior and individual comfort demands of
each occupant of a building.
Integration between demands of end-user and
building is the ultimate in the intelligent building
579
Zeiler W., van Houten R., Boxem G., van der Velden J., Wortel W., Haan J., Noom P., Kamhuis R., Hommelberg M. and Broekhuizen H. (2009).
MULTI-AGENTS FOR ENERGY EFFICIENT COMFORT - Agents for the Energy Infrastructure of the Built Environment: Flexergy.
In Proceedings of the International Conference on Agents and Artificial Intelligence, pages 579-586
DOI: 10.5220/0001534705790586
Copyright
c
 SciTePress

concept. “Connecting” the end-user to a building is
complex. User-connectivity, the combination of
usability and user interface together, is studied and
developed further. Information and communication
technology connects people and helps them to
communicate with the building ( Clements-Croome,
1997).
A new generation of building process control
systems is being developed based on agent
technology. Experiences from earlier work proved
that the traditional approaches are not capable to
coope with the increasing complexity of multi-
agents structures for energy infrastructures within
the total built environment. New approaches are
needed.
In section 2 there is a description of experiments
with agent technology. The change in the future
energy infrastructure is described in section 3. Based
on the experiences gained in earlier projects a
methodology is developed to cope with the
necessary changes, which is presented in section 4.
The overall resulting framework to structure and to
implement agent technology in a integral way within
the built environment is presented in section 5.
2 EARLIER EXPERIMENTS
Previous work by Akkermans (2002) showed that
agent technology makes it possible to integrate
occupants’ behaviour. Multi-agent systems provide
the essential technology for this information
infrastructure to connect the end-user to the building
systems (Akkermans, 2002);
- large numbers of actors are able to interact, in
competition or in cooperation
- local agents focus on local interests and
negotiate with more global agents
- implementation of distributed decision making
by the negotiation processes between the different
local or more global oriented agents
- communication between actors is minimized to
generic information exchange between agents.
To cope with different users and their different
needs system wide information by agents is the
basis. The different agents dynamically and
continuously exchange information and negotiate
with each other to get the best conditions for their
representative. Through this mechanism there is an
exchange of information about needs and supplies
throughout the whole system. Only in this way the
system can cope with the different users and their
different needs. In two projects, SMART (Smart
Multi Agent internet Technology) (Jelsma et al.
2002) and IIGO ( Intelligent Internet mediated
control in the built environment) (Kamphuis et al.
2005) this technology was developed and tested.
A different type of technology to incorporate
user behavior, Forgiving Technology, was
developed in another project, EBOB, Energy
Efficient Behavior in Office Building (Claeson-
Jonsson, 2005). EBOB investigated new combined
technical and socio-economic solutions to make
energy efficient behavior natural, easy and
intuitively understandable for the end-users of
refurbished and new offices. Control scenarios for
the HVAC (Heat, Ventilation and Air-conditioning)
systems were derived by analyzing occupant’s
behavior and its effects on comfort and energy use.
The EBOB project is an European 5th
framework program project with eleven partners
from five countries. EBOB ran from 2002 until
2005. The field test was held at Kropman’s office at
(Grundelius et al., 2004).
The techniques used within SMART/IIGO and
EBOB made it possible to use the user
representation and combine it with optimalization
techniques. The representation of end-users was
realized by developing an individual voting system.
End-users were represented in the design by
Fanger’s comfort model (Fanger, 1970). This
comfortmodel predicts user’s evaluations of the
indoor climate in buildings. This Predicted Mean
Vote model (PMV) is the basis of the indoor climate
standards in Europe ISO 7730-2005 and America,
ASHRAE Standard 55-2004. This model includes
thermo physiological properties of humans, such as
sweat production and heat resistance of the skin.
Based on what average people consider comfortable,
the Predicted Mean Value (PMV) is translated into a
percentage of people dissatisfied (PPD). Using the
model of Fanger, the percentage of dissatisfied users
can be predicted for a given set of comfort
parameters. The voting system allowed every user in
a thermal zone to enter his vote (warmer/colder)
within a voting period (e.g. one hour) while seeing
the aggregated voting of other users in his zone at
the moment of voting (Jelsma et al. 2002).
The users comfort needs dominate this control
strategy. The control strategy is based on the
description of the user behaviour and implemented
in a BMS (Building Management System). This
BMS was extended with an external real-time
information system to improve energy and comfort
control. A learning curve is built from the user
voting behavior. Responses of the user are
interpreted differently depending on the overall
ICAART 2009 - International Conference on Agents and Artificial Intelligence
580

trend of the comfort level in the building. Overall
voting behavior as a function of the time of day is
included in determining the action of the local
comfort aspect controllers. Within this system the
persistent use of user information is a leading
strategy.
By starting from the human perspective and using
available and new technology (including IT, smart
control, user interfacing) this dominate user strategy
was achieved. Figure 1 shows an overview of the
agent system as part of the building management
system with the individual voting behaviour. All the
agents are communicating with other agents,
representing rooms or the floors of the building.
Also there are agents representing the information
about the weather forecast and the central process
control of the air-handlings unit. The relation
between the individual demands and the changing
external weather conditions is also shown in figure
1.
Figure 1: Individual adjustments and different energy
demands for each office room shown on the screen of the
BMS (Hommelberg 2005).
3 FUTURE CHANGE IN THE
ENERGY INFRASTRUCTURE
Due to the change in buildings, equipment and out
door climate (growing cooling demand), there is a
strongly growing demand for electricity instead of
heating. As electricity has a completely different
character as energy form compared to heat, a
completely different strategy is needed to optimize
the energy infrastructure of the built environment.
Heat can be stored rather easily and efficient, were
as electricity can only be stored in a limited amount
in expensive and complex devices. As result of this
change the focus of the design process will have to
change too. The flexibility of electricity is almost
zero as storage possibilities are rather limited and
have relatively poor performances.
Therefore it is important to look at energy
reduction especially for this growing electricity
demand of a building.
Electricity is traditionally generated in large
central plants and distributed throughout the country.
During the last decades this is changing. More and
more decentralized electricity production is done by
means of wind turbines, combined heat power units
and photovoltaic systems. This will change bit by bit
the whole distribution system from a strict top down
system to a bottom-up system in which user can
supply electricity in to the distribution grid on
different levels.
For the users this means that instead of only using
centralized electricity production, users can use
different electricity sources by their own or others.
The making of the built environment and its
necessary energy supply grid has become complex.
A flexible energy infrastructure in and between
buildings is needed to optimize the combination of
decentralized power generation, use of sustainable
energy source on building level and traditional
centralized energy supply. The energy flows of heat,
cold and electricity have to be optimized together.
Preservation of energy resources, occupant comfort
and environmental impact limitation are the key
issues of modern and sustainable built environment.
Although the experimental field tests applied with
the multi agent process control systems proved
successful and led to a stable it also proved that a
more integral approach was needed to further
optimize comfort and energy use in a building. All
the energy flows such as heat, cold, electricity have
to be optimized in connection to each other not only
on the level of a specific building but on the level
the built environment. For such a complex system
approach the bottom-up approach starting from
building segments is not enough for building a
integral multi-agent process control. A new design
approach is needed to structure the different layers
and different functional defined tasks for agents in
such an integral multi-agent process control system.
To achieve this design knowledge plays an essential
role.
New integral design approaches and design support
are needed. To achieve this design knowledge plays
an essential role. As stated by Gruber et al. (2006)
the spectrum of expressiveness and degree of
knowledge ranges from simple lists of terms or
MULTI-AGENTS FOR ENERGY EFFICIENT COMFORT - Agents for the Energy Infrastructure of the Built
Environment: Flexergy
581

vocabularies over taxonomies and database schemas
up to ontologies (Guarino 1998, Corcho et al. 2003).
4 METHODOLOGY
4.1 Ontology for Design
‘Ontology’ in philosophy means theory of existence
in the broadest sense. It tries to explain what is being
and how the world is configured by introducing a
system of critical categories to account things and
their intrinsic relations (Kitamura 2006). In the
knowledge engineering community an ontology is
viewed as a shared conceptualization of a domain
that is commonly agreed to by all parties. It is
defined as ‘a specification of a conceptualization’(
Gruber 1993). “Conceptualization’ refers to the
understanding of the concepts that can exist or do
exist in a specific domain or a community. A
representation of the shared knowledge in a specific
domain that has been commonly agreed to refers to
the ‘specification’ of a conceptualization ( Dillon et
al. 2008).
An ontology aims to capture the conceptual
structures in a domain by describing facts assumed
to be always true by the community of users.
Ontology is the agreed understanding of the ‘being’
of knowledge: consensus regarding the interpretation
of the concepts and the conceptual understanding of
a domain (Dillon et al. 2008)
Ontology is generally considered to provide
definitions for the vocabulary used to represent
knowledge. The ontology role is to reflect a
community’s consensus on a useful way to
conceptualize a particular domain (Aparrício et al.
2005). Ontology building deals with modeling a
domain of the world with shareable knowledge
structures (Geller et.al 2004).
Based on observations from literature, Uschold
(1998) identified three main categories of uses for
ontologies (see Figure 1; for further details and
examples see Uschold & Gruninger (1996)):
- Communication between people. Here, an
unambiguous but informal ontology may be
suffcient.
- Inter-operability among systems achieved by
translating between different modelling methods,
paradigms, languages and software tools; here,
the ontology is used as an interchange format (see
Figure 2).
- Systems engineering benefits: in particular,
- Re-usability: the ontology is the basis for a
formal encoding of the important entities,
attributes, processes and their inter-relationships
in the domain of interest. This formal
representation may be (or become so by
automatic translation) a re-usable and/or shared
component in a software system
5.
- Knowledge acquisition: speed and reliability
may be increased
- Reliability: a formal representation also
makes possible the automation of consistency
checking resulting in more reliable software.
- Specification: the ontology can assist the
process of identifying requirements and defining
a specification for an IT system (knowledge
based, or otherwise).
Figure 2: Interchange format example. This illustrates the
use of an ontology as an interchange format to integrate
different software tools (Uschold 1998).
Ontologies are formal conceptualizations not
made l’art pour l’art, but to help achieve a goal or
task by an actor. That task involves knowledge-
intensive reasoning to understand the world not just
static, but to serve practical purposes of action by
the actor in his world (Akkermans 2008).
4.2 Prescriptive Design Method:
Integral Design
Designing is a creative activity using several kinds
of knowledge. The quality of design relies heavily
on knowledge applied in the design processes
(Kitamura 2006).
Design knowledge sharing is expected to
drastically improve the design process. For example,
in activities related to design review, an explicit
ICAART 2009 - International Conference on Agents and Artificial Intelligence
582

description of the designer’s intentions helps other
people to understand the original design more
effectively. Even designers themselves can gain
deeper insights into the designs themselves
(Kitamura et al. 2004).
More than two decades of knowledge
engineering have shown that there are recurring
patterns or stereotypes in the structuring and use of
knowledge as an instrument in tasks that involve
reasoning and computing. One of these recurring
knowledge stereotypes are problem-solving
methods; heuristic and stereotypical in the sense that
they do not guarantee to solve a given knowledge-
intensive problem in general. These problem-solving
methods do have demonstrated pragmatic value in
solving typical or common cases of knowledge-
intensive task that can, moreover, be reused in many
different situations (Akkermans 2008). There is a
strong analogy between the problem-solving
methods and prescriptive design methods.
The design process has been a topic of design
research resulting in large numbers of models and
theories of design, yet there is no consensus ( Sim
and Duffy 2003). What is well common among most
models of engineering design processes is the
depiction of the design process as consisting of
conceptual distinct phases or stages of activities that
transform the design from a set of requirements to a
final design solution (Sim and Duffy 2003).
Engineering design can thus be viewed as an
articulate process composed of phases, where each
phase represents a combinatorial action on the parts
the composite object is constituted of (Colombo et.al
2007).
To develop our required model of design
support, an existing model from the mechanical
engineering domain was extended: Methodical
Design by van den Kroonenberg (de Boer 1989,
Blessing 1994) into an Integral Design methodology
(Zeiler 2000). The Integral design process can be
described at the conceptual level as a chain of
activities which starts with an abstract problem and
which results in a solution. The design activity can
be divided into four phases: clarification of the task,
conceptual design, embodiment design and detail
design (Camelo et al. 2007). The original methodical
design process is extended from three to four main
phases, in which different levels of functional
hierarchical abstraction, stages can be distinguished.
Hierarchical abstraction implies the decomposition
of information into levels of increasing detail, where
each level is used to define the entities in the level
above. In this sense each level forms the abstract
primitives of the level above. The contents of the
layers are based on the technical vocabularies in use,
technology-based layers or levels. Each layer
represents an abstraction of the levels below.
4.3 Functional Decomposition Model
In order to survey solutions, engineers classify
solutions based on various features. This
classification provides a mean to decompose
complex design tasks into manageable problems. An
important decomposition is based on functions.
While there is no common understanding of what a
function is, people share the idea that functional
knowledge is tightly related to design intention
(Kitamura et.al 2004). Functions, as a concept,
seems to derive from the designer’s intention and it
has no clear, unified, objective, and widely accepted
definition (Umeda and Tomiyama 1997). Still when
designers speak about the ‘function’ held by an
object or by one of its components, they can speak
about it because they have sufficient knowledge for
associating functions to a suitable object structure (
Colombo et al. 2007). Starting by formulating the
need, the program of demands is developed and
transformed into functions to fulfil. Functions can be
regarded as what a design is supposed to fulfil: the
intended behaviour of the object.
During the design process, and depending on the
focus of the designer, functions exist at the different
levels of abstraction. The functional decomposition
is carried out hierarchically so that the structure is
partitioned into sets of functional subsystems.
This functional decomposition provides the
means for decomposing complex design tasks into
problems of manageable size. This functional
decomposition is hierarchically so that the structure
is partitioned into sets of functional subsystems.
Decomposition is done until simple building
components remain whose design is a relatively easy
task. So functions play a crucial role in a design
process, because the results of the design depend
entirely on the decomposition of the function (
Umeda and Tomiyama 1997).
The concept of hierarchical functional
abstraction levels leads to a structure of different
sets of functions for cooling, heating, lighting,
power supply and ventilation, see figure 3.
It represents the orderings principle: abstraction
levels, main functions and sub functions.
MULTI-AGENTS FOR ENERGY EFFICIENT COMFORT - Agents for the Energy Infrastructure of the Built
Environment: Flexergy
583

Built Environment
Building
Floor
Room
Workplace
Human
Need
Supply
Storage
Distribution
Storage
Central
Distribution
Generation
Cooling
Heating
Ventilation
Elektricity
Light
Abstraction levels; Main functions; Subfunctions;
Figure 3: Abstraction levels, main functions and
subfunctions.
4.4 Orientations in the Design
Knowledge Model
Designing takes place in an environment that
influences the process, it is contextually situated (de
Vries 1994, Dorst and Hendriks 2001). The context
of the model of designing is defined by a “world
view”. The model consists of four worlds: the real
world R, the symbolic world S, the conceptual world
C and the specification world M. Thus, the four
levels of aspect abstraction in the descriptive model
of design are:
1. Information Level; knowledge-oriented,
representing the "conceptual world. This level deals
with the knowledge of the systems by experts. One
of the essential ideas behind this is that human
intelligence has the capability of search and the
possibility to redirect search. This information
processing is based on prior design knowledge. One
of the major problems in modeling design
knowledge is in finding an appropriate set of
concepts that the knowledge should refer to, or in
more fashionable terms; an ontology (Alberts 1993).
2. Process Level; process oriented, representing
the "symbolic world". This level deals with physical
variables, parameters and processes. The set of
processes collectively determines the functionality
of the variables that represent the device properties.
Modelling at the functional level involves the
derivation of an abstract description of a product
purely in terms of its functionality. This abstraction
reduces the complexity of engineering design to the
specification of the product’s desired functionality.
3. Component Level; device orientation,
representing the "real world. This level describes the
hierarchical decomposition of the model in terms of
functional components and is domain dependent.
Generic components represent behaviors that are
known to be physically possible to realize. They are
generic in the sense that each component stands for a
range of alternative realizations. This also implies
that the generic components still have to be given
their actual shape.
4. Part Level; parametric orientation,
representing "the specification world". This level
describes the actual shape and specific parameters of
the parts of which the components exist. Relevant
technical or physical limitations manifest themselves
in the values of a specific set of parameters
belonging to the generic components. These
parameters are used to get a rough impression, at the
current level of abstraction, of the consequences of
certain design choices for the final result.
The four levels of aspect abstraction in the
descriptive model of design can be related to the
hierarchical levels used within the functional
decomposition, see figure 4. The ontology can now
be used to generate new possibilities for a flexible
process control energy infrastructure in and between
buildings to optimize the combination of
decentralized power generation, use of sustainable
energy source on building level and traditional
centralized energy supply.
Need
Programme of
demands
Design
Brief
User
Solution
Built
environment
Building
Floor
Room
Workingplace
Information level
Process level
Component
level
Part level
Figure 4: Hieracrchical abstraction levels by functional
decomposition.
This makes it possible to integrate in a flexible way
the energy flows connected to heating, cooling,
ventilation, lighting and power demand, within a
building and between buildings and the built
environment. This leads to flexibility of energy
exchange between different energy demands and
sustainable energy supply on the different levels of
abstraction in the built environment.
ICAART 2009 - International Conference on Agents and Artificial Intelligence
584

5 DISCUSSION
The research tradition related to “ontology” within
Computer Science especially Information and
Knowledge Systems is now about twenty years old.
The field of ontology has been extremely successful
in strictly or in terms of socio-economic or industrial
usefulness (Akkermans 2008). Geller et al. (2004)
give a thumbnail historical perspective on ontology
and its challenges can be found. Looking at
conferences held during the last years the ontology
penetration rate is high and has become a
cornerstone in areas such as semantic web, database,
engineering, business and medicine. However
looking more closely, it can be noticed that these
events are attended by companies and enterprises
only in minimal part and the feeling remains that the
‘fuss’ about ontology is mainly at the level of
research and its surrounding niches (Borgo and
Lesmo 2008). Industry seems not to recognize the
value of applying insights of ontology research.
In engineering practice an important critical
issue is the distance, in terms of intuitiveness and
perceived complexity, between the engineered
ontology and its intended user of application
community (Akkermans 2008).
Our proposed model for supporting the design
process by the Integral design method process frame
work is simular to the functional approach
developed by Stone and Wood. They propose a
systematic placement of components into a
hierarchical ontology (Bryant Arnold et al. 2007),
using the functionality of components as a natural
framework upon which such abstractions can be
built. So they use more a kind of bottom-up
approach compared to our integral approach.
6 CONCLUSIONS AND FURTHER
RESEARCH
A functional decomposition framework based on
hierarchical abstraction is proposed as a theoretical
basis for design of the agent process control for the
building, its building services systems and its energy
infrastructure. We think that the proposed
framework supports Multi-Agent technology in
optimizing the energy infrastructure within the built
environment.
Using two types of knowledge levels modesls
(Uschold 1998), ontology and problem solving
model, led to an approach in which the
characteristics of the combined models offer an
added value to design within the built environment
domain.
A new integral design methodology has been
developed and is used to develop a flexible concept
for further development and implementation of the
new design and control strategy for new energy
infrastructures for the built environment:
Flex(ible)(en)ergy
The TU/e (Technische Universiteit Eindhoven)
together with Kropman, Installect and ECN (Energy
research Centre Netherlands) work on research for
user based preference indoor climate control
technology. Central in this approach is the user focus
of the integral building design process which makes
it possible to integrate sustainable energy more
easily in the energy infrastructure and reduce energy
consumption by tuning demand and supply of the
energy needed to fulfill the comfort demand of the
occupants building.
Taking the user as starting point a new Multi-Agents
framework is defined to optimize the process control
within a flexible sustainable energy infrastructure;
Flex(ible)(en)ergy.
At the moment this approach is implemented within
the Flexergy project. This project started in 2007 and
will continue till 2010.
ACKNOWLEDGEMENTS
Kropman and ECN were partners in the SMART and
IIGO project, which was partly financial supported
by SenterNovem. Kropman bv and the foundation
“Stichting Promotie Installatietechniek (PIT)”
support the research. Flexergy project is financial
supported by SenterNovem, project partners are
Technische Universiteit Eindhoven, ECN and
Installect.
REFERENCES
Akkermans, H., 2002, Being Smart In Information
Processing. Technological And Social Challenges And
Oppertunities, Proceedings IFIP IIP2002
Akkermans, H., 2008, The Business of Ontology calls for
a Formal Pragmatics, Invited talk 3rd workshop on
Formal Ontologies Meet Industry, FOMI 2008, 5-6
juni Torino, Italy.
http://e3value.few.vu.nl/docs/bibtex/pdf/HansA2008.p
df
Alberts, L.K., 1993, YMIR: An Ontology for Engineering
Design, PhD-thesis, University of Twente, Enschede.
Alley, R. et.al., 2007, Climate Change 2007: The Physical
Science Basis Summery for Policymakers,
MULTI-AGENTS FOR ENERGY EFFICIENT COMFORT - Agents for the Energy Infrastructure of the Built
Environment: Flexergy
585

Intergovernmental Panel on Climate Change, Paris,
France
Aparriíco, A.S., Farias, O.L., Santos N. dos, 2005,
proceedings Australasian Ontology Workshop (AOW
2005), Sydney. Conferences in Research and Practice
in Information Technology (CRPIT),Vol.58.
ASHRAE 55, 2004, ANSI/ASHRAE Standard 55-2004,
Thermal Environmental Conditions for Human
Occupancy, American Society of Heating,
Refrigerating and Air-conditioning Engineering,
Atlanta.
Blessing, L.T.M.,1994, A process-based approach to
computer supported engineering design, PhD-thesis
Universiteit Twente.
Boer, S.J. de, 1989, Decision methods and techniques in
methodical engineering, PhD-Thesis Universiteit
Twente
Borgo, S., Lesmo, L., 2008, The Attractiveness of
Foundational ontologies in Industry, Formal
Ontologies Meet Industry, IOS Press.
Bryant Arnold, C.R., Stone, R.B., Greer, J.L., McAdams,
D.A., Kurtoglu, T., Campbell, M.I., 2007, A
Function-based component ontology for system
design, Proceedings International conference on
engineering design, ICED’07, 28-31 August, Paris,
France
Camelo, D., Mulet, E., Vidal, R., 2000, Function and
behaviour representation for supporting flexible
exploration and generalisation in a functional model
for conceptual design, 2007, Proceedings International
conference on engineering design, ICED’07, 28-31
August, Paris, France
Claeson-Jonsson, C., 2005, Final publishable report,
Energy Efficient Behaviour in Office Buildings.
Clements-Croome, T.D.J, 1997,What do we mean by
intelligent buildings?, Automation in Construction 6
(1997) 395-400
Colombo, G., Mosca, A., Sartori, F., 2007, Towards the
design of intelligent CAD systems: An ontological
approach, Advanced Engineering Informatics 21
(2007) 153-168.
Corcho, O., Fernández-López, M. And Gómez-Pérez A.,
2003, Methodologies, tools and languages for building
ontologies. Where is their meeting point?, Data &
Knowledge Engineering 46(1), pp. 41-64
Dillon, T., Chang, E., Hadzic, M., Wongthongham, P.,
2008, proceedings Fifth Asia-Pacific Conference on
Conceptual Modelling (APCCM 2008), Wollongong,
NSW, Australia, January 2008.
Dorst, K., Hendriks, D., 2001, The Role of the Design
Context: In Practice and in Design Methodology,
Proceedings 5th International Design Thinking
Research Symposium “Designing in Context” , 18-20
December 2001, Delft
Fanger, P. O. , 1970, Thermal comfort Analysis and
Applications in Environmental Engineering, McGraw-
Hill, London,New York, ISDN 0-07-0199154
Geller, J., Perl, Y., Lee, J., 2004, Ontology Challenges: A
Thumbnail Historical perspective, Editorial,
Knowledge and Information Systems (2004) 6: 375-
379.
Gruber, T.R., 1993, A translation approach to portable
ontologies, Knowledge Acquisition, 5(2):199-220
Grundelius, M., Lundberg, S., Brissman, J., Stalpers-
Croez,e I., Wortel, W., Bakker, L., Soethout, L.,
Scholten, J.E., Woksepp, S., 2004, EBOB Total result
of WP2, Forgiving Technologies, Prototype 2
description, D22 2.5 and D17 WP2, 2004-06-30,
www.ebob-pro.com/reports.htm
Guarino, N., 1998, Formal Ontology and Information
Systems, proceedings of FOIS’98, Trento, Italy, 6-8
june 1998, Amsterdam IOS Press, pp. 3-15,
http://www.loa-cnr.it/Papers/FOIS98.pdf
Hommelberg, M.P.F, 2005,
Careful Buildings, software
agents voor een gebouwbeheersysteem, Master
thesis Building Services TU/e, January 2005 ( dutch)
ISO 7730, 2006, Gematigde thermische binnencondities.
Bepaling van de PMV- en PPD-waarde en specificatie
van de voorwaarden voor thermische behaaglijkheid,
(in Dutch) guideline, NEN-EN-ISO 7730
Jelsma, J., Kets, A., Kamphuis, I.G. and Worte,l W., 2002,
SMART work package 2002 final report; SMART
field test: experience of users and technical aspects,
research report ECN-C—02-094, ECN, Petten.
Kamphuis, I.G., Warmer, C.J., Jong, M.J.M., Wortel, W.,
2005, IIGO: Intelligent Internet mediated control in
the built environment: Description of a large-scale
experiment in a utility building setting, ECN rapport
ECN-C—05-084, October 2005.
Kitamura, Y., 2006, Roles of ontologies of engineering
artifacts for design knowledge modeling, In Proc. Of
the 5th International Seminar and Workshop
Engineering design in Integrated product Development
(EDIProD 2006), 21-23 September 2006, Gronów,
Poland, pp.59-69.
Kitamura, Y., Kashiwase, M., Fuse, M., Mizoguchi, R..
2004, De ployment of an ontological framework of
functional design knowledge, Advanced Engineering
Informatics 18 (2004) 115-127.
Sim, S.K., Duffy, A.H.B., 2003, Towards an ontology of
generic engineering design activities, Res Eng Design
14 (2003) 200-22.3
Umeda, Y. and Tomiyama T., 1997, Functional
Reasoning in Design, IEEE Expert, march-april, pp.
42-48
Uschold, M. and Gruninger, M. (1996): Ontologies:
Principles, methods and applications', The Knowledge
Engineering Review 11(2). (Also available as AIAI-
TR-191 from AIAI, The University of Edinburgh.
Vries, T.J.A. de, 1994, Conceptual design of controlled
electro-mechanical systems, a modeling perspective,
PhD thesis Twente university, Enschede, ISBN:90-
9006876-7
Zeiler, W., 2000, Integral Design and Building
Performance Simulation, proceedings IBPSA
symposium Building performance simulation in the
Benelux, 27 October 2000, Eindhoven University of
Technology
ICAART 2009 - International Conference on Agents and Artificial Intelligence
586