Multi-robot Cooperation Based on Learning Social Models
Rosiery da Silva Maia, Luiz Marcos Garcia Gonçalves · 2012
IDeM-MRS is a cooperation formalism for multi-robot systems to efficiently solve tasks in indoors environments. This formalism is based on learning social models of real individuals and it coordinates the group allowing knowledge assimilation and accommodation through experiences exchange. Some issues are particularly investigated, such as the realist multi-robot environment representation (involving the global mission, the tasks that belong to this mission and the active robots) and a task selection form based on theories and social learning approaches, which allow the cooperative and efficient execution by robots. IDeM-MRS can be used for different types of missions, from the simplest to more complex ones. The experiments and results validate the efficiency of this formalism compared to the traditional empirical model.