Biologically inspired stochastic hybrid control of multi-robot systems

This paper describes a biologically inspired approach to the modeling and control of multi-agent populations composed by a large number of agents. Individual agents are modeled based on a deterministic Hybrid Automata endowed with input events and continuous-valued outputs. The complexity of population modeling is avoided by assuming a stochastic approach, under which the agents distribution over the state space is modeled. This is based on a Stochastic Hybrid Automaton, which results from inputting a stochastic event sequence to the individual model. The dynamics of the state probability density functions is determined and the results applied to the mission control of a simulated robotic population.