Decentralized square-root information Kalman filtering based on J-orthogonal transformations

Abstract

The paper addresses the problem of decentralization of the measurement data processing based on matrix J-orthogonal transformations in the square-root information Kalman filter. J-orthogonal transformations are applied in a decentralized filter at the stage of communication and assimilation. The results of numerical experiments confirm the efficiency of the proposed solution. The developed algorithm can be used for measurement data processing in decentralized distributed computing networks.