论文信息 - A multiagent interaction paradigm for physiological process control

A multiagent interaction paradigm for physiological process control

Multiagent systems are powerful and flexible tools for controlling complex phenomena. The complexity of a phenomenon can be tackled in such a way that each agent embeds the controller for a portion of the phenomenon [3]. In this perspective, the interaction among the agents results in a complex controller for the whole phenomenon. The actions the agents undertake to control their portions of the phenomenon may conflict, as a result of the "overlapping" of the controlled portions; hence a mediated interaction is needed. A class of complex phenomena that present several difficulties in their satisfactory modeling and controlling is the class of physiological processes.We illustrate the negotiation paradigm of a general regulating multiagent architecture called anthropic agency.

[1] E. Salzsieder,et al. KADIS: model-aided education in type I diabetes. Karlsburg Diabetes Management System. , 1994, Computer methods and programs in biomedicine.

[2] Julie A. Adams,et al. Multiagent Systems: A Modern Approach to Distributed Artificial Intelligence , 2001, AI Mag..

[3] NICHOLAS R. JENNINGS,et al. An agent-based approach for building complex software systems , 2001, CACM.

[4] Leonid Sheremetov,et al. Weiss, Gerhard. Multiagent Systems a Modern Approach to Distributed Artificial Intelligence , 2009 .

[5] C. Cobelli,et al. A physiological simulation model of the glucose-insulin system , 1999, Proceedings of the First Joint BMES/EMBS Conference. 1999 IEEE Engineering in Medicine and Biology 21st Annual Conference and the 1999 Annual Fall Meeting of the Biomedical Engineering Society (Cat. N.

[6] N. R. Jennings,et al. To appear in: Int Journal of Group Decision and Negotiation GDN2000 Keynote Paper Automated Negotiation: Prospects, Methods and Challenges , 2022 .