Achieving consensus in self-organizing multi agent systems for smart microgrids computing in the presence of interval uncertainty

Self-organizing multi agent systems, namely cooperative agent networks employing decentralized computing paradigms based on dynamic populations of mutually coupled oscillators, are assuming a major role in supporting the large-scale deployment of smart microgrids (SMGs). The adoption of these architectures would allow the agents to compute the main global variables characterizing the SMG operation without the need for a central fusion center. Thanks to this feature, all of the basic SMG control and monitoring functions could be processed according to a totally decentralized/non-hierarchical computing paradigm. Anyway, effectiveness of these architectures on real environment should face several issues not fully explored in the literature. Amongst these, the effect of data uncertainty has been recognized as one of the most critical issue to address. In trying and fixing this problem, this paper proposes new formalizations of the decentralized consensus protocols based on the use of Interval Arithmetic. The application of this reliable consensus protocols to decentralized SMG computing is explained in detail and several numerical results are presented and discussed in order to assess the effectiveness of the proposed approach.