An Adapative Mulit-Agent Architecture for Collaborative Problem Solving in Command Post of the Future

The intelligent data processing tools of the command posts of the future must be able to take full advantage of the huge amount of available information by having a mechanism for the fusion of disparate types of data. The fusion of data can yield some information which may not be obtainable when the data are interpreted separately. The proposed work focuses on providing a general architecture for the fusion of real time heterogeneous data. This architecture is general enough to be used with different fusion algorithms. It allows for the addition of new fusion algorithms and new data types. The architecture also allows for adaptation to real time changes in the data. Two popular areas of Al are utilized: intelligent software agents and genetic algorithms. The independent nature of intelligent agents in a multi-agent environment allows them to work independently and in parallel on heterogeneous databases. The results can then be integrated through a collaborative schema. Genetic algorithms are utilized because of their efficiency in searching large data spaces and their adaptability to real time change. To increase human comprehensibility, advanced graphical visualization techniques are used to display the outputs of various modules of the system.