Fault tolerant multi-sensor data fusion for autonomous navigation in future civil aviation operations

The problem addressed in this paper is that of fault-tolerant multi-sensor data-fusion for flight navigation during approach and landing of a civil jetliner. The work falls within the scope of a large collaborative project on future single pilot operations issues. The proposed methodology is based on a multi-layer hierarchical architecture which can integrate several heterogeneous information sources. The study is restricted to three classical information sources, namely, Inertial Reference System (IRS), Global Positioning System (GPS) and Instrument Landing System (ILS). The system is designed to meet the specified navigation performance requirements in terms of validity, accuracy, continuity and on-board implementation issues. The methodology takes advantage of recently developed set-membership tools, and paves the way toward a new and simple tuning of global merging filter.