Robust Fault Detection and Isolation algorithms for TVC systems: An experimental test

Abstract

This paper addresses the development and hardware testing of Fault Detection and Isolation (FDI) algorithms for Thrust Vector Control (TVC) systems driven by Electro-Mechanical Actuators (EMAs). Reusable rockets can primarily benefit from such algorithms because assessing any TVC degradation during a mission is essential to timely trigger recovery actions. Furthermore, monitoring TVC sensor can also evidence the presence of degradation across multiple flights. We propose a suitable FDI architecture and test it on a dedicated Hardware-In-the-Loop (HIL). The goal is to ensure that nominal and faulty conditions can be correctly discerned. This study employs a model-based FDI approach that relies on a nullspace computation, which already achieved a high TRL and a relevant track of record. The algorithms are synthesized to be robust against disturbance, like friction and vehicle-induced loads. To this purpose, a detailed modeling of the system, as well as fault detectability and isolability analyses are provided.