부산대학교 도서관

This paper presents a novel distributed state-estimation framework for satellite formations. Existing consensus-based distributed filters have high communication re-quirements that make them unsuitable for resource-constrained small-spacecraft formations. We present a low-communication distributed filtering scheme targeted at small-satellite formations that are limited by inter-satellite communication chan-nel capacity. We assume a chief-deputy topology, similar to the proposed NASA Helioswarm mission. Our novel state-estimation framework is based on decoupling the chief state estimation filter from the deputies and using two separate filter architectures: one chief-specific that accounts for intermittent access to inertial position measurements, and a second deputy-specific one that only incorporates relative range measurements between the spacecraft, having the chief state as an external model input. We introduce a consider-covariance technique to account for the chief's state uncertainty in the deputy's filter. We demonstrate that the proposed distributed filter architecture requires significantly less communication than prior consensus-based methods in the literature while achieving comparable accuracy.