Awareness-based Couplings of Intelligent Agents and Other

Advanced Coupling Concepts for M&S

Tuncer Ören

and Levent Yilmaz

School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada

Dept. of Computer Science and Software Engineering, Auburn University, Auburn, AL, U.S.A.

Keywords: Agent Coupling, Awareness-based Coupling, Perception-based Coupling, Anticipation-based Coupling,

Introspection-based Coupling, Agent Monitored Coupling, Deliberation-based Couplings.

Abstract: This study is a sequel of a recent publication where the coupling concepts as well as the advantages of

coupling of declarative models were clarified. In this article, the basic concepts of simulation model

coupling is reviewed. Advanced input concepts, including context-mediated perception, are elucidated.

Synergy of agents and simulation is revised. Awareness-based couplings of intelligent agents are explained.

Other advanced coupling concepts clarified include: deliberation-based coupling, introspection-based

coupling, anticipation-based coupling, and model/real-system coupling.

1 INTRODUCTION

System theoretic formulation of coupling of

component models was first introduced in the

seminal work of Wymore (1967) on mathematical

theory of systems engineering. The concept allows

the formulation of input-output relationships of

component-based systems as well as systems of

systems. The first application of system coupling in

simulation modeling was published by Ören (1971)

for GEST ‒ the first system-theory-based declarative

simulation modeling language (for continuous and

for piece-wise continuous system modeling). In a

recent article, a systematic and comprehensive view

as well as advantages of hard couplings of

declarative simulation models was elaborated by

Ören (2014). The concept of coupling has been

successfully applied to discrete event system

specification (DEVS) by Zeigler (1984) and many

other researchers who use and/or contribute to the

DEVS formalism. What has been expressed and

achieved until now can be labelled as hard coupling

as well as conventional coupling.

In this article, our aim is to elaborate on soft

coupling and extend the coupling concept for use in

modeling and simulation with autonomous and

quasi-autonomous intelligent systems. More

specifically, promising possibilities of awareness-

based coupling of intelligent software agents are

pointed out. We will examine soft couplings under

two dimensions: proximity and activation. Proximity

will be classified in terms of the cognitive,

conceptual, and physical spaces, and to define

activation, we will review the “input” concept.

In this article the following is done:

(1) Fundamental concepts of model coupling are

briefly reviewed in Section 2. Model coupling is

basically input-output relationship of atomic or

resultant of already coupled models. As such,

model coupling represents communications of

component models.

(2) To explore several types of communications

hence several types of inputs as bases for

generalized coupling concepts, non-conventional

types of inputs to simulation models, such as (i)

cognitive inputs (perception, introspection, and

anticipation), (ii) sensations and especially (iii)

chemical signals (pheromones) are covered, in

Section 3. Perception, introspection, and

anticipation (of both external and internal

knowledge) form also basis for machine

awareness. In information processing, the

information traces would have similar functions

as pheromones in nature.

(3) Highlights and importance of the several types of

synergy of software agents and simulation

(which lead to agent-directed simulation) are

noted in Section 4.

(4) In Section 5, three types of awareness-based

couplings, i.e., perception-based, introspection-

Ören T. and Yilmaz L..

Awareness-based Couplings of Intelligent Agents and Other Advanced Coupling Concepts for M&S.

DOI: 10.5220/0005473700030012

In Proceedings of the 5th International Conference on Simulation and Modeling Methodologies, Technologies and Applications (SIMULTECH-2015),

pages 3-12

ISBN: 978-989-758-120-5

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

based, and anticipation-based couplings of

software agents as well as other advanced types

of couplings are elaborated on.

(5) In Section 6, two types of coupling of

simulations and real systems are explained.

(6) Finally, in Section 7, we express our conclusion

about this very important and promising research

area and outline our future activities.

Over 90 terms denoting several types of

couplings of simulation models are listed in an

appendix.

2 MODEL COUPLING: BASICS

Some of the papers on simulation model coupling

published by one of the authors are (Ören, 1971,

1984, 2014). Since in the last one, the coupling

concept is explained in detail, here we cover only

very basic concepts to provide the foundation for

more advanced types of couplings. Figure 1, adopted

from (Ören, 2014) depicts a typical coupling of

component models.

Figure 1: A coupled model Z (adopted from Ören, 2014).

Figure 2 (also from Ören, 2014) represents a

template for model coupling. It consists of four

parts: (1) Externals, i.e., coupled model’s (or

resultant model’s) list of external inputs and outputs;

(2) component models where a list of names of the

component models to be coupled is given; (3)

external coupling; and (4) internal coupling. Details

of the coupling specification of Z, given in Figure 1

was covered in (Ören, 2014).

Coupled Model model-identifier

Externals -- input and output variables of the

coupled model

Inputs -- list of external inputs

Outputs -- list of external outputs

End Externals

Component Models

-- List of names of component models

End Component Models

External Coupling

Inputs

-- equivalencing external inputs to internal inputs

-- for every external input specify

-- to which input of which component

model(s) it is connected to

End Inputs

Outputs

-- equivalencing external outputs to internal

outputs

-- for every external output specify

-- to which output of which component model

is connected to

End Outputs

End External Coupling

Internal Coupling

-- for every component model

-- for every input variable (which is not connected

an external input) specify

-- from which output variable of which component

model the values are provided

End Internal Coupling

End Coupled Model model-identifier

Figure 2: Template for model coupling (adopted from

Ören, 2014).

3 ADVANCED INPUT CONCEPTS

To be able to explore different types of input/output

relationships of (atomic as well as already coupled

component models, it is imperative to consider all

SIMULTECH2015-5thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

kinds of inputs. In an article, Ören (2001a)

dichotomized inputs as exogenous inputs (i.e.,

externally generated inputs) and endogenous inputs

(i.e., internally generated inputs) and elaborated on

each category. Some other publications on types of

inputs are (Ören and Yilmaz, 2004; and Yilmaz and

Ören, 2009). An updated version of the classification

of the types of inputs is given in Tables 1 and 2.

Table 1 summarizes externally generated inputs or

exogenous inputs. As shown in Table 1, perceived

inputs are generated outside of a system; however,

the system needs to actively discern (or recognize)

them.

Table 1: Types of exogenous (externally generated) inputs

(Adopted from Ören and Yilmaz, 2009).

Exogenous Inputs

Mode Type

Imposed

input

(Passive

acceptance of

exogenous

input)

Access to input

- Direct input, coupling, argument

passing, knowledge in a common area

(blackboard), message passing,

broadcasting (to all, to a fixed or

varying group, to an entity)

Nature of input

- Information

Data, facts, events, goals

- Sensation

(Converted sensory data (Table 3)

from analog to digital – single or

multi sensor – sensor fusion)

Perceived

input

(Active

perception of

exogenous

input)

Perception process includes

Noticing, recognition, decoding,

selection (filtering), regulation

Nature of input

- Interpreted sensory input data

(Table 3) and selected events

- Infochemicals (Table 4) (chemical

messages/chemical messengers

for chemical communication)

--sources:

---animate

---inanimate

- Infotraces (traces of information

transactions among:

--interconnected infohabitants of

---Internet of things

---Users of media and

search engines

Table 2 represents internally generated inputs or

endogenous inputs. There are two categories of

internally generated inputs: intropection, i.e.,

perception of internal knowledge (or realization of

lack of some knowledge) and anticipated inputs.

Types of sensations mentioned in Table 1 are

elaborated on in Table 3.

Table 2: Types of endogenous (internally generated)

inputs (Adopted from Ören and Yilmaz, 2009).

Endogenous Input

Mode Type

Perceived

endogenous

input

Introspection

Perceived (cogitated) internal facts,

events; or realization of lack of them

Anticipated /

deliberated

input

Anticipation

Anticipated facts and/or events

(behaviorally anticipatory systems)

Deliberation

Deliberation of past facts and/or

events (deliberative systems)

Generation

Generation of goals, questions and

hypotheses by:

- Expectation-driven reasoning

(Forward reasoning, or

(Bottom up reasoning, or

(Data-driven reasoning)

- Model-driven reasoning

Table 3: Types of sensations (Adopted from Ören and

Yilmaz, 2009).

Type

of stimulus

Type of perception

light

- vision (visual perception): visible light

vision, ultraviolet vision, infrared vision

sound

- hearing (auditory sensing): audible /

infrasonic / ultrasonic sound (medical

ultrasonography, fathometry)

chemical

- (gas sensing / detection):

smell (smoke / CO2 / humidity sensor)

- (solid, fluid sensing): taste, microanalysis

heat

- heat sensing

magnetism

- magnetism sensing:

geomagnetism / thermo-magnetism

sensing, electrical field sensing

touch

- sensing surface characteristics

motion

- acceleration sensing

vibration

- vibration sensing: seismic sensor

Chemical signals, or infochemicals are especially

important in plants and animals (and even in

inanimate entities) to assure several types of

chemical communications. Table 4 categorizes some

chemical signals (infochemicals or chemical

messages / chemical messengers for chemical

communication).

Awareness-basedCouplingsofIntelligentAgentsandOtherAdvancedCouplingConceptsforM&S

Table 4: Types of infochemicals for chemical

communication.

Infochemicals

(Chemical messengers for chemical communication)

Nature

Animate

Hormones (Interactions are within a living

organism between different organs or

tissues)

Semio-

chem-

icals

Interactions are:

Intra-

specific

(same

species)

Pheromones

(In early literature

Echtohormones)

Inter-

specific

(different

species)

Allelochemics

(allomones,

antimones,

kairomones,

synomones)

Inani-

mate

Apneumones

Infochemicals are chemical messages generated

by animate or inanimate beings to influence the

physiology or behavior of part of self, same or

different species. They are very interesting,

challenging, and pragmatically important.

Especially, in biologically-inspired computing, they

are very important and inspiring.

“Hormones are chemical messengers that are

secreted directly into the blood, which carries them

to organs and tissues of the body to exert their

functions. There are many types of hormones that

act on different aspects of bodily functions and

processes. Some of these include: development and

growth, metabolism of food items, sexual function

and reproductive growth and health, cognitive

function and mood, and maintenance of body

temperature and thirst” (News-medical). In

modeling physiological systems, functional coupling

of the components organs can be expressed as time

varying couplings.

Semiochemicals are chemicals emitted by an

organism to influence the physiology or behavior of

an organism of the same or a different species. They

include pheromones and allelochemics. Applications

of semiochemicals include insect pest control with

limited or no contamination of environment. Some

characteristics of semiochemicals are categorized in

Table 5.

Pheromones are intraspecific behavior altering

chemical agents. Based on their functions, there are:

aggregation pheromones, alarm pheromones, caste-

regulating pheromones, releaser pheromones, sex

pheromones, and trail marking pheromones.

Allelochemics are chemical messengers “produced

by a living organism that exerts a detrimental

physiological effect on individuals of another

species when released into the environment” (OD).

Apneumones are chemicals “released by a non-

living substance that is beneficial to the receiver”

(Capinera, 2008, p. 230).

Table 5: Types of some allelochemics, based on their

affection characteristics (Adopted from Barrows (2011, p.

102).

Signal to the

benefit of

receiver

yes

sender

yes

Synomones

(e.g., floral sent,

pollinator)

Allomones

(defence secretion,

repellant; e.g., venom of

snake/person)

Kairomones

(e.g., a parasite

seeking a host)

Antimones

(e.g., chemicals of a

pathogene/host)

4 SYNERGY OF AGENTS AND

SIMULATION

Most often, simulation of agent systems or

simulation with agent-based models is considered.

However, to see and get the full benefits, synergy of

simulation and agents which is called Agent-

Directed Simulation (ADS) is very important

(Yilmaz, Ören, 2009). As shown in Table 6, synergy

of simulation and agents consists of contributions of

simulation for agents and contributions of agents for

simulation.

Table 6: Types of agent-directed simulation.

Types of simulation

Synergy of simulation and agents

Agent-directed simulation (ADS)

Contributions of

simulation to

agents

Agent simulation

- Simulation of agent systems

or simulation with agent-

based models.

Contributions of agents to

simulation

Agents as

support

facilities

Agent-supported simulation

- Agent support for

user/system interfaces

- Agents to enhance cognitive

capabilities of modeling and

simulation systems.

Agents as

monitoring

facilities

Agent-monitored simulation

- Includes model behavior

generation.

- Agent-monitored coupling.

SIMULTECH2015-5thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

Simulation for agents (short form for ‘‘contributions

of simulation to agents’’) is called agent simulation.

Most common type of agent simulation is simulation

of agent systems (or simulation of any system

modeled by software agents or agent-based models).

Agent simulation is also called ‘‘agent-based

simulation’’ by those who do not take into account

contributions of agents to simulation. Already many

applications of agent simulation exist in engineering,

human, and social dynamics, civilian as well as

military applications. Sometimes the term “multi-

agent” systems is also used.

Contributions of agents for simulation consist of

two categories of possibilities: (1) use of agents as a

support facility to enable computer assistance in

problem solving and/or enhancing cognitive

capabilities of simulation systems and (2) use of

agents for simulation run-time activities.

As support facility, agents can support front-end

user/system interface functions, such as problem

specification or back-end user-system interface

functions, such as data compression, explanation,

problem and/or solution documentation, and solution

selection. Agents can also enhance cognitive

capabilities of modeling and simulation systems by

providing understanding and multi-understanding

abilities.

Use of agents for simulation run-time activities

includes model behavior generation as well as other

activities such as agent-monitored model update and

agent-monitored coupling.

While dynamic composability, interoperation,

and run-time extensibility in agent simulation is

highly sought, contemporary coupling solutions

often lack mechanisms for (dynamic) selection and

assembly, as well as meaningful run-time

interoperation among agents. In particular, they are

limited in dealing with (1) dynamically evolving

content (i.e., data, model) needs of existing federated

agent applications and (2) run-time inclusion of new

agents into a federated system with their own

encoding standards and behavioral constraints.

Besides, existing interoperation strategies are not

transparent to the actual simulation infrastructure.

Agent-monitored simulation also covers agent-

monitored coupling where one or more agents

examine interaction protocol between multiple

agents to facilitate mediation, brokering, and

matchmaking services (

Yilmaz and Paspuletti, 2005).

Agents can provide various functions while

monitoring and analyzing the coupling between

agents. Administration is the process of managing

the information exchange needs that exist between

agents.

Administration involves the overall

information management process for the agent

architecture. Location, discovery, and retrieval of

content are critical components of administration.

Management involves identifying, clarifying,

defining and standardizing the meaning of content.

Alignment ensures that the data to be exchanged

exist in the participating agents as an information

entity or that the necessary information can be

derived from the available services published by the

agents. Transformation is the technical process of

aggregation and/or disaggregation of the information

entities of the embedding systems to match the

information exchange requirements.

5 AWARENESS-BASED

COUPLINGS

Several types of cognitive input-output relationships

are possible for intelligent agents. The input-output

relations in the cognitive space include perception-

based couplings, introspection-based couplings, and

anticipation-based (Ören and Yilmaz, 2012)

couplings. In general they can be categorized as

awareness-based couplings.

Introspection-based coupling is part of self-

awareness of agents. The awareness-based couplings

will further be enriched by some other advanced

concepts such as variable couplings (including time-

varying couplings) and nested couplings as clarified

by Ören (1971, 2014). These concepts lead to

context-sensitive coupling.

To characterize soft coupling in the conceptual

and physical spaces, we propose mechanisms that

serve as filters to selectively allow agents to

indirectly interact with others via the physical and

conceptual space. In Tables 1-4, the input concept is

extended with the signalling construct to facilitate

agents to communicate with each other through

infochemicals induced into the environment.

Diffusion, aggregation, and evaporation of signals

will enable time-varying coupling fields that emerge

as agents interact in the physical and conceptual

space. Similar to infochemicals, information traces

can be used in information systems.

5.1 Context-mediated Perception

Chemical communication and coordination provide

apt metaphors for devising advanced perception-

based coupling strategies between agents. Dicke and

Sabelis (1989) clarified infochemical terminology. A

recent and more comprehensive clarification is given

by Barrows (2011).

Awareness-basedCouplingsofIntelligentAgentsandOtherAdvancedCouplingConceptsforM&S

An infochemic is a chemical that conveys

information between two interacting entities,

triggering in the receiver a behavioural or

physiological response that is adaptive to either one

of the interacting agents or to both (Dicke and

Sabelis, 1989).

A pheromone provides a context-mediated

coupling mechanism by facilitating interaction

between organisms of the same species. The

interaction benefits the organism that originates the

pheromone, to the receiver, or to both.

On the other hand, allelochemics enable interaction

between two entities belonging to different species

(Table 4). Allelochemics are divided into

synomones, allomones, kairomones, and antimones.

In the case of synomones, the chemical released

by the source organism results in behavior that is

adaptively favorable to both the source (sender) and

the receiver organisms. The pollination process

serves as a good example that benefits both the

plants and the insects. Allomone is a chemical that is

pertinent to an organism such that when it contacts

with an individual of another organisms, it evokes in

the receiver a behaviour that is favorable to the

originating organism. For instance, some emit toxins

to deter herbivores to keep them away. A kairomone

is a chemical that is pertinent to the source organism

and that when it contacts with a second organisms, it

triggers behavior in the target organism that is

adaptively favorable to it, but not to the source

organism. The release of chemical cues that attract

predators is an example for kairomone. Antimone is

a chemical substance which may be detrimental to

the host of if its emitter. Barrows (2011) provides a

wealth of information about allelochemics and

pheromones.

Using the above metaphors, environment-

mediated, indirect, and perception-based couplings

can be defined in terms of their function in the

interactions between agents. Among such functional

couplings are alarm pheromones, aggregation and

spacing pheromones, and diffusion (gossip)

pheromones. The perception of a pheromone may

trigger (releaser effect) a reactive in motion by the

first perception (primer effect). Similarly,

allelochemics can be used as a metaphor to define

coupling functions such as enemy-avoidance and

foraging kairomones. Tuning various properties of

the infochemical-inspired signals can moderate

coupling mechanisms between agents. The

regulation can include volatility of the signals,

stability in the context (environment), and rate of

diffusion.

The benefits of perception-based couplings

based on such context-mediated signals include their

(1) effectiveness in the absence of direct coupling

between agents that have different interfaces, or the

presence of large number of diverse agents with

coupling needs, (2) ability to provide time-coded

signals and hence generating temporal effects, (3)

capability to remain in an environment for an

extended period of time, and (4) ability to avoid

close proximity between coupled agents.

5.2 Deliberation-based Coupling

Agents with internal models of the environment as

well of as peer agents can use deliberative

mechanisms to perceive the intention of other agents

and/or interpret their behavior. Interaction decisions

are made selectively based on the goals, tasks, and

activities that cohere together to achieve the desired

high-level objectives. Agents use the current state

and perceptions to deliberate and generate awareness

models prior to initiating interaction decisions.

Deliberation-based couplings may also be

important in simulation with holonic agents. Holonic

agent simulation or holon simulation, in short, is an

important type of agent simulation where agents

represent holons. “Holonic systems are excellent

candidates to conceive, model, control, and manage

dynamically organizing cooperative systems” (Ören,

2001b). An important type of cooperation is co-

opetition, i.e., limited cooperation of otherwise

competitive entities. “A holonic system is composed

of autonomous entities (called holons) that can

deliberately reduce their autonomy, when need arise,

to collectively achieve a goal. A holonic agent is a

multi-agent system where each agent (called a

holon) acts with deliberately reduced autonomy to

assure harmony in its cooperation in order to

collectively achieve a common goal” (Ören, 2001b).

5.3 Introspection and Anticipation-

based Coupling

Agents with internal models of themselves (or a self-

observing module) can interpret their own

knowledge processing activities as well as perceive

internal facts, events; or realization of lack of them.

Hence, these internally generated knowledge can be

used in some couplings as inputs. Relevant part of

the coupling would then be introspection-based

coupling. Ören and Yilmaz (2004) elaborated on

behavioral anticipation in agent simulation.

Knowledge generated by a behaviorally anticipatory

entity can be an input to another relevant entity. The

SIMULTECH2015-5thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

associated coupling is then an anticipation-based

coupling.

6 SIMULATION/REAL-SYSTEM

COUPLING

Table 7 represents possibilities for the relationships

of the operations of the simulation and the real

system. Most often, operations of simulation and

real world system are not connected. This type of

simulation is called standalone simulation. However,

there are two cases where operations of simulation

system and the real world can be closely related.

This is the case of integrated simulation where

simulation model can receive input directly from the

real system. In this case one can consider

simulation/real-world coupling.

Simulation model and real-world can be coupled for

two reasons: To enrich or to support real system’s

operation. As clarified in Table 6, when

simulation/real-world coupling is done to enrich

real system’s operation, the system of interest and

the simulation program operate simultaneously. Two

possible uses are: (1) online diagnostics (or

simulation-based diagnostics) and (2) simulation-

based augmented/enhanced reality operation (for

training to gain/enhance motor skills and related

decision making skills).

When simulation/real-world coupling is done to

support real system’s operation, the system of

interest and the simulation program operate

alternately to provide predictive displays. These

parallel experiments while system is running would

permit to use simulation to decide whether or not

some or all decision variables should be changed.

Such couplings are especially useful for systems

characterized by non-linear interactions among

diverse agents that exhibit emergent behavior, which

may be very different from what the initial

conditions of these systems would suggest.

Traditional simulation techniques that rely on

accurate knowledge of these conditions typically fail

in these cases. The Symbiotic Adaptive

Multisimulation (SAMS) strategy (Mitchell and

Yilmaz, 2009; Yilmaz and Mitchell, 2009) enables

robust decision making in real-time for these types

of problems. Rather than relying on a single

authoritative model, SAMS explores an ensemble of

plausible models, which are individually flawed but

collectively provide more insight than would be

possible otherwise.

Table 7: Types of simulation with respect to the

connectivity (or coupling) of operations of simulation and

real system (Adopted from Ören, 2009, p. 10).

Type of

connectivity

Type of simulation

Operations of the simulation and the real system are

not connected Standalone simulation

interwoven (integrated simulation)

To enrich real

system’s

operation

(The system of interest and the

simulation program operate

simultaneously)

- online diagnostics (or

simulation-based diagnostics)

- simulation-based

augmented/enhanced reality

operation (for training to

gain/enhance motor skills

and

related decision making

skills)

To support

real system’s

operation

(The system of interest and the

simulation program operate

alternately to provide

predictive displays)

- parallel experiments while

system is running

The insights derived from the model ensemble

are then used to improve the performance of the

coupled system under study. Likewise, as the system

behavior unfolds, observations of emerging

conditions can be used to improve exploration of the

model ensemble. In essence, a useful co-evolution

between the physical system and SAMS occurs.

Self-simulation via SAMS provides a framework to

generate anticipatory effects to explore system

behavior. Through simulation/real-world coupling, a

self-simulating system maintains accurate and

consistent models of itself and the environment.

7 CONCLUSIONS AND FUTURE

WORK

The concepts discussed in this article may enrich

agent-based modeling and may also be inspirational

for advanced coupling in simulation in general and

in agent-directed simulation in particular. We are

planning two courses of action:

(1) The generalization of the multi-model concept

(Yilmaz and Ören, 2005) to multi-model agents.

Afterwards their couplings will be elaborated on.

(2) Implementation of some of the advanced

coupling concepts, especially awareness-based

coupling concept for agents with emotion

Awareness-basedCouplingsofIntelligentAgentsandOtherAdvancedCouplingConceptsforM&S

understanding abilities for emotional intelligence

simulation.

This article ‒with its 90+ types of model

couplings‒ may also be a rich reference for

modeling and simulation body of knowledge.

REFERENCES

Barrows, E.M. (2011). Animal behavior desk reference: A

dictionary of animal behavior, ecology, and evolution.

Taylor & Francis Boca Raton, FL, USA.

Capinera, J.L. (Ed.) (2008). Encyclopedia of Entomology,

vol. 4. Springer Science & Business Media.

Dicke M. and M. W. Sabelis (1989). Infochemical

terminology: Based on cost-benefit analysis rather

than origin of compounds? Functional Ecology, vol. 2,

no. 2, pp. 131–139.

Mitchell B. and L. Yilmaz (2009). Symbiotic Adaptive

Multisimulation: An Autonomic Simulation

Framework for Real-time Decision Support Under

Uncertainty. ACM Transactions in Modeling and

Computer Simulation. 19:1, article 2, 31 pages.

(News-medical). http://www.news-medical.net /health/

What-are-Hormones.aspx.

OD – Oxford Dictionaries http://www.oxforddictionaries.

com/definition/english/allelochemical.

Ören, T. (1971). GEST: A combined digital simulation

language for large-scale systems. Proceedings of the

Tokyo 1971 AICA (Association Internationale pour le

Calcul Analogique) Symposium on Simulation of

Complex Systems, Tokyo, Japan, September 3-7, pp.

B-1/1 - B-1/4.

Ören, T.I. (1984). GEST - A Modelling and Simulation

Language Based on System Theoretic Concepts. In:

Simulation and Model-Based Methodologies: An

Integrative View, T.I. Ören, B.P. Zeigler, M.S. Elzas

(eds.). Springer-Verlag, Heidelberg, Germany, pp.

281-335.

Ören, T. (2001a - Invited paper). Software Agents for

Experimental Design in Advanced Simulation

Environments. In: S.M. Ermakov, Yu.N. Kashtanov,

and V.Melas (eds.) Proc. of the 4th St. Petersburg

Workshop on Simulation, June 18-23, 2001, pp. 89-95.

Ören, T.I. (2001b). Advances in Computer and

Information Sciences: From Abacus to Holonic

Agents. Special Issue on Artificial Intelligence of

Elektrik (Turkish Journal of Electrical Engineering

and Computer Sciences - Published by TUBITAK -

Turkish Science and Technical Council), 9:1, 63-70.

Ören, T.I. (2009). Modeling and Simulation: A

Comprehensive and Integrative View. In L. Yilmaz

and T.I. Ören (eds.). Agent-Directed Simulation and

Systems Engineering. Wiley Series in Systems

Engineering and Management, Wiley-Berlin,

Germany, pp. 3-36.

Ören, T. (2014–Invited review paper). Coupling Concepts

for Simulation: A Systematic and Comprehensive

View and Advantages with Declarative Models.

International Journal of Modeling, Simulation, and

Scientific Computing (IJMSSC), vol. 5, issue 2 (June):

pp. 1‒17 (This article is a revised version of the

invited paper presented at the Conference of the

Chinese Academy of Engineering (CAE), Wuxi, China,

Oct. 12, 2013.).

Ören, T.I. and L. Yilmaz (2004). Behavioral Anticipation

in Agent Simulation, Proceedings of WSC 2004 -

Winter Simulation Conference, pp. 801-806.

Washington, D.C., December 5-8, 2004.

Ören, T. and L. Yilmaz (2012). Agent-monitored

anticipatory multisimulation: A systems engineering

approach for threat-management training. Proceedings

of EMSS'12 – 24th European Modeling and

Simulation Symposium, F. Breitenecker, A. Bruzzone,

E. Jimenez, F. Longo, Y. Merkuryev, B. Sokolov

(Eds.), September 19-21, 2012, Vienna, Austria, pp.

277.282. ISBN 978-88-97999-01-0 (Paperback), ISBN

978-88-97999-09-6.

Wymore, A.W. (1967). A mathematical theory of systems

engineering: the elements. Wiley.

Yilmaz, L. and B. Mitchell (2009). Autonomic

Introspective Simulation Systems. In: Agent-Directed

Simulation and Systems Engineering. Levent Yilmaz

and Tuncer Ören (eds.) Wiley-Berlin, Germany. pp.

37-72.

Yilmaz, L. and T. Ören (2005). Discrete-Event

Multimodels and their Agent-Supported Activation

and Update. In Proceedings of the Agent-Directed

Simulation Symposium of the Spring Simulation

Multiconference (SMC’05), pp. 63-70, San Diego, CA,

April 2005, pp. 63-70.

Yilmaz, L. and T. Ören (2009). Agent-directed

Simulation. In: Agent-Directed Simulation and

Systems Engineering. Levent Yilmaz and Tuncer Ören

(eds.) Wiley-Berlin, Germany. pp. 111-143.

Yilmaz L. and S. Paspuletti (2005). Toward a Meta-Level

Framework for Agent-supported Interoperation of

Defense Simulations," Journal of Defense Modeling

and Simulation. Vol. 2, no 3, pp. 161-175.

Zeigler, B.P. (1984). Multifacetted Modeling and Discrete

Event Simulation. Academic Press, London; Orlando.

APPENDIX

Terms denoting types of model couplings (over 90

types)

--A--

Affective coupling

Agent coupling

Agent-aided coupling

Agent-controlled coupling

Agent-monitored coupling

Allelochemic coupling

Allomone coupling

SIMULTECH2015-5thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications

Anticipation-based coupling

Antimone coupling

Awareness-based coupling

--B--

Basic coupling

Broadcasted coupling

--C--

Cascade coupling

Cognitive coupling

Common coupling

Computer-aided coupling

Conjunctive coupling

Content coupling

Context-insensitive coupling

Context-mediated coupling

Control coupling

Conventional coupling

Coupling

Coupling of variable component model

Coupling with variable connection

--D--

Data coupling

Data-structured coupling

Decoupling

Deliberation-based coupling

DEVS coupling

Direct coupling

Disjunctive coupling

DNA-based coupling

Dynamic coupling

Dynamic federate-coupling

Dynamic model-coupling

--E--

Emotional coupling

Environment-mediated coupling

Essentially cascade coupling

External coupling

Feedback coupling

Functional coupling

--G--

Generalization coupling

GEST coupling

Graphic coupling

--H--

Hard coupling

Hierarchical coupling

Holonic agent coupling

--I--

Indirect coupling

Infochemical coupling

Informational coupling

Input/output coupling

Intermodular coupling

Internal coupling

Introspection-based coupling

--K--

Kairomone coupling

--L--

Limiting coupling

Logical coupling

Loose coupling

Loose temporal coupling

Low coupling

--M--

Miscoupling

Mixed coupling

Model coupling

Model/real-system coupling

Multi-level dynamic coupling

Multi-model agent coupling

Multi-model coupling

--N--

Nested coupling

Nonlinear coupling

Nonlinear statistical coupling

--P--

Perception-based coupling

Persistent coupling

Pheromone coupling

Pure feedback coupling

--R--

Resultant coupling

Runtime coupling

--S--

Sequential coupling

Semantic coupling

Singular coupling

Soft coupling

Specialization coupling

Awareness-basedCouplingsofIntelligentAgentsandOtherAdvancedCouplingConceptsforM&S

Subtype coupling

Supertype coupling

Stamp coupling

State coupling

State-dependent coupling

State-independent coupling

Static coupling

Structural coupling

Subclass coupling

Synomone coupling

System coupling

--T--

Targeted coupling

Temporal coupling

Tight coupling

Time-dependent coupling

Time-invariant coupling

Time-varying coupling

Topological coupling

--V--

Variable coupling

Volatile coupling

SIMULTECH2015-5thInternationalConferenceonSimulationandModelingMethodologies,Technologiesand

Applications