Air Quality Management Using a Multi-Agent System.

Management of urban atmospheric pollution necessitates advanced modeling and information processing techniques. The design of the prototype system DNEMO, which is based on a distributed adaptive problem-solving approach is the focus of the research reported in this paper. Issues covered in the paper relate to the distributed nature of this environmental problem, handling noise and uncertainty in monitoring data, achieving graceful degradation of performance and system robustness, and adaptation of system performance to long-term evolution of the monitored phenomena. The research reported can be applicable to a broad class of environmental monitoring applications, since the problems addressed are common to many environmental problems. [ABSTRACT FROM AUTHOR]