Robust FDI applied to thruster faults of a satellite system

This paper presents a practical solution to the problem of robust fault detection and isolation (FDI) for faults affecting the thrusters of a satellite system Mars Express (MEX). A non-linear model of MEX is used and the system is considered in the presence of measurement errors, uncertainty and disturbance. A robust approach to FDI residual generator design is described to realise a diagnostic scheme for the satellite system, focused on a “thruster modulation” problem. The approach taken is based on both state estimation of an accurate linear model for the satellite system and unknown input de-coupling to achieve robust FDI in the presence of severe dynamic uncertainty during main engine deployment. A brief description of the satellite structure is outlined in order to highlight the main sources of uncertainty and disturbance affecting the monitored system. The satellite nominal and faulty conditions are simulated by means of an industry standard non-linear model implementation in the MATLAB® and SIMULINK® environments. A comparison with the existing fault diagnosis scheme implemented on board the MEX is also reported. Suitable performance criteria are introduced together with Monte Carlo analysis for both design tuning and assessing the robustness and performance evaluation of the complete FDI design strategy.