Novel Sliding Innovation Filter Inspired Fault Detection for Hydrofoil Attitude Control Systems

In this paper, a novel approach for detecting anomalies in the non-linear fully-submerged hydrofoil attitude control system (HACS) is proposed, even in the presence of time-varying disturbances. To address the trade-off between robustness against disturbances and optimality in terms of estimation error, the extended sliding innovation filter (SIF) is employed as a state estimator for the target non-linear HACS. By utilizing a switching gain with a sliding boundary layer, the SIF inherently possesses a degree of robustness to estimation issues that may involve fault conditions or factors of disturbances. A residual framework is subsequently established to achieve state tracking and comparison. The residual evaluation for the fault detection (FD) scheme is then easily conducted using statistical methods such as the modified Z-Score and the peak signal-to-noise ratio (PSNR). Finally, the effectiveness of the developed FD strategy is substantiated through experiments conducted on a hardware-in-loop (HIL) platform. Comparative analysis with state-of-the-art robust UKF algorithms reveals the impressive fault detection proficiency of the proposed strategy.