A Hyperparameter-Based Approach for Consensus Under Uncertainties

Abstract

This paper addresses the problem of information consensus in a team of networked agents with uncertain local estimates described by parameterized distributions in the exponential family. In particular, the method utilizes the concepts of pseudo-measurements and conjugacy of probability distributions to achieve a steady-state estimate consistent with a Bayesian fusion of each agent's local knowledge, without requiring complex channel filters or being limited to normally distributed uncertainties. It is shown that this algorithm, termed hyperparameter consensus, is adaptable to any local uncertainty distribution in the exponential family, and will converge to a Bayesian fusion of local estimates over arbitrary communication networks so long as they are known and strongly connected.