A decision network based framework for multiagent coalition formation

Novel systems allocating teams of humans and unmanned heterogeneous vehicles are necessary for future applications. An intelligent framework is presented that reasons over a library of coalition formation algorithms to select the most appropriate algorithm(s) to apply to complex missions. The framework is based on decision networks to handle uncertainties in dynamic environments. A group of features is used to identify the most suitable algorithm(s). The proposed framework uses principal component analysis to extract the most significant features that are crucial for making decisions. A technique based on link analysis calculates the utility values for each feature-value pair and algorithm in the library. Experimental results demonstrate that the presented framework accurately chooses the most appropriate coalition formation algorithm(s) based on multiple specified mission criteria and requirements.