Analysis of dynamic conflicts by techniques of artificial intelligence
Abstract
Dynamic conflicts exhibit differentiel game characteristics and their analysis by any method which disregards this feature may be, by definition, futile. Unfortunately, realistic conflicts may have an intricate information structure and a complex hierarchy which don't fit in the classical differential game formulation. Moreover, in many cases even well formulated differential games are not solvable. In the recent years great progress has been made in artificial intelligence techniques, put in evidence by successfull applications in scientifique modelling, automated engineering design processes as well as for fuzzy and intelligent control systems. This progress has raised hopes that artificial intelligence methods can be of help also in solving complex dynamic conflicts. This scientific report outlines a feasible option which combines artificial intelligence techniques with concepts of differential game theory for attaining such an objective. A research effort in this direction has a great potential of success, but it requires a well planned and coordinated collaboration of qualified scientist in both disciplines.