What are Agents?
Agent-based systems are of increasing importance. They are regarded as a new paradigm enabling an important step forward in empirical sciences, technology and theory.
This new paradigm has an enormous influence on model building and simulation. Novel design patterns and architectural structures are opening up new possibilities. Furthermore, new areas of application are coming into the foreground.
Cooperative agents can be equipped with different attributes which however do not all appear to be necessary or useful in every application. The following are of particular importance:
Every agent is characterised by autonomous behaviour. This consists in the first instance of the so-called inherent dynamism that an intelligent agent displays autonomously without input from the environment.
In addition there is the induced dynamism which describes how the intelligent agent reacts in response to inputs from the environment.
Strategies that guide the behaviour of an agent when it is pursuing its goals are of particular importance here.
Individual World View
Every agent has its own model of the external world that surrounds it. This so-called conceptual model describes how the intelligent agent sees the world. The conceptual model will generally be incomplete. In some circumstances it is even incorrect.
The manner in which the intelligent agent builds up its model of the world on the basis of the information it receives from its environment is of particular interest.
Communicative and Cooperative Capacity
Intelligent agents can exchange information with their environment and with other intelligent agents. By means of the possibility of communication an intelligent agent must obtain information about its environment which enables it to build up its own world model.
Moreover the possibility of communication with other intelligent agents is the precondition of common action in pursuit of a goal.
As the term Intelligent Agent indicates, behavioural possibilities such as the capacity to learn, logical deduction or construction of an environmental model in order to find one’s bearings in an unknown environment and suchlike are required.
Intelligent agents are sometimes required to display spatial mobility. Spatial mobility is in fact a characteristic of numerous intelligent agents. On the other hand there are elements in the real world that are not spatially mobile but which one would nonetheless wish to assign to intelligent agents. For this reason spatial mobility in this study will not be considered as one of the necessary attributes of intelligent agents.
Modelling and Simulation
The Modelling and Simulation of real systems consisting of intelligent agents that cooperate with each other has recently emerged as an important field of research. Systems of this kind can be found in the empirical sciences, in technology and in mathematical theories. In this context two areas are of particular interest:
Strategies and Decentral Control
The object is to develop inidividual strategies that individual agents pursue and that ensure that a common goal can be achieved even without central regulation.
Cooperation between intelligent agents can produce a stable system that displays new global behaviour on the next higher level of abstraction. The task is to explain this global behaviour on the basis of the individual attributes of the intelligent agents and of their interactions.