Meta-synthesis information fusion for hybrid diagnostics of space avionics

Abstract

This article describes meta-synthesis information fusion as a novel mode in integrated avionic management systems. With deeper space explorations, avionic health management systems require more perfectly integrated data and information because of the increased spacecraft functionality and the complex software. Meta-synthesis information fusion allows for a more accurate picture of the state of the avionics and therefore allows for better decision making. This study uses a meta-synthesis information fusion application for the hybrid diagnostics, which is a very important part of integrated health management systems. For the meta-synthesis information fusion, specific approaches such as probability theory, neural networks and the Dempster–Shafer evidence theory are used to construct the hybrid diagnostics model. Through this novel meta-synthesis information fusion mode, efficiency as a whole, from input to output is realized and dynamic, real-time diagnostics is achieved. A numerical example is given, which demonstrates the application of the hybrid diagnostics to a radar indicator. By analyzing the feasibility and the pragmatic utility of the hybrid diagnostics meta-synthesis information fusion, the advantages of this mode are shown.