Intellectual Development Model for Multi-Robot Systems

摘要

We propose the IDeM-MRS learning formalism to be used by a group of robots for solving practical tasks in indoor environments. The formalism is inspired on the theory of social learning models for human beings that is traditionally developed in Psychology and Education fields. Our model can be used for coordination of the group, as for, allowing assimilation and accommodation of knowledge through experience exchange. Besides explaining the theoretical model itself, we formalize the mathematics involved with it in a very simple and straightforward fashion. Some issues are especially investigated such as the realistic representation of the multi-robot environment involving the global mission, the tasks belonging to the mission and the active set of robots. A way for task selection is proposed based on social learning theories and approaches that allow cooperative and efficient execution of tasks by robots. To this end, IDeM-MRS can be used in different types of missions varying from simple to complex. Experiments and results validate the efficiency of the formalism compared to a traditional empirical model.