OTKA által támogatott kutatási projekt.

Nyilvántartási szám: T 038081

The research is aimed at the elaboration, implementation and application of novel methodologies for solving problems by model identification and reconstruction using simulation with AI controlled agents in a wide range of application fields. Within that – starting from a model based on an initial working hypothesis – the aim is the determination of model structures, operational mechanisms, functions representing the interactions and parameters describing the models more adequately. The research work deals with the solution of problems of multidisciplinary "soft" as well as confrontative and cooperative systems. To achieve these ends efficient knowledge representation structures, inferencing, decision and intervening algorithms are to be elaborated, implemented and verified by practical applications. The research is to be undertaken by the scientific staff of the International McLeod Institute of Simulation Sciences Hungarian Center – hosted at the Budapest University of Technology and Economics, Faculty of Economic and Social Sciences, Department for Information and Knowledge Management – operating as a member of the international network of excellence and lead by Professor András Jávor Ph.D.,D.Sc., director of the Center. In the research work there is a close scientific cooperation with other centers of the International McLeod Institute of Simulation Sciences engaged in activities related to our research work. The results achieved are to be applied as a decision support tool to solve problems relevant for the society.

Main tasks: Investigation of efficient simulation procedures of identification and reconstruction of multidisciplinary and basically "soft" systems by AI controlled agents with regard to optimal knowledge base structures describing the model structures and operational mechanisms, efficient search algorithms and ensuring a wide range of model modifications. Implementation and testing of intelligent agent (demon) structures. Investigation and application of methodologies based on the interaction of cooperative and confrontative intelligent agents for simulating economic processes and conflict situations. Investigation of the structure and application of mobile entities equipped with knowledge bases and inference engines. Investigation of the application of intelligent agents at models characterized by continuous variables.

Results to be attained: The theoretical elaboration, implementation and experimental investigation of intelligent agents enabling the identification of multidisciplinary "soft" as well as cooperative and confrontative systems. Solution of practical problems in case of multidisciplinary "soft" systems, cooperative as well as confrontative problems by means of simulation using AI controlled agents. Elaboration of the structure of intelligent active mobile entities and their application in simulation. The application of intelligent agents to simulation models characterized by continuous variables.