Dynamic Modeling and Fault Diagnosis Technologies for Aircraft Landing Gear Systems: A Comprehensive Review
Abstract
Aircraft landing gear systems are characterized by strong nonlinearity, multi-physics coupling, and significant uncertainty. This paper reviews representative modeling approaches, including multibody dynamics, rigid–flexible coupling modeling, multi-domain unified modeling, co-simulation, and tire–runway coupled modeling. Furthermore, the applicability of fault tree analysis, bond graph-based diagnosis, data-driven diagnosis, and model-data fusion diagnosis is examined. The results indicate that physics-based models provide strong interpretability and support airworthiness verification; however, they involve high modeling costs and perform poorly in real time under complex operating environments and parameter uncertainties. Purely data-driven methods excel at extracting nonlinear features but are constrained by fault sample scarcity and the long-tail distribution of fault modes. Fusion diagnosis, digital twin technology, and hardware-in-the-loop (HIL) validation are regarded as promising solutions for balancing accuracy, interpretability, and engineering feasibility. Future research should focus on high-fidelity reduced-order modeling, intelligent diagnosis under limited samples, lifecycle-oriented digital twins, and airworthiness-oriented validation frameworks to support the design, health monitoring, and predictive maintenance of large civil aircraft landing gear systems.