A Hybrid Health Monitoring Approach for Aircraft Flight Control Systems With System-Level Degradation

This article proposes a novel hybrid health monitoring approach to monitor the system-level degradation of aircraft flight control systems (FCSs). The idea derives from observing the nonlinear hysteresis phenomenon in FCS. First, a health FCS model is developed to implement the nonlinear degradation signal extraction of FCS by building an adaptive-network-based fuzzy inference system, a data-driven method. Subsequently, the jump Markov autoregressive exogenous (JMARX) system with time delays is adopted to establish the FCS system-level degradation model. An expectation maximum-convex optimization algorithm is innovatively proposed to identify the model parameters. After that, three health indicators associated with the degradation model parameters are utilized for FCS system-level health monitoring. Finally, a practical flight experiment is conducted by a civil aircraft. The obtained experimental data is used to validate the effectiveness of the proposed FCS monitoring approach. The model of the time-delay JMARX system gets good modeling evaluation results on mean absolute error, standard deviation, etc. Besides, each of the three health indicators shows a clear FCS degradation tendency, which indicates that the proposed method successfully extracts the degradation information and monitors the health states of FCS. This approach is potential for practical and effective engineering applications in the aviation industry.