HEXmorph: Fault tolerance against single and dual rotor failure using geometric morphing on hexacopter

Abstract

Rotor failure in hexacopters with alternating rotor configurations often results in propulsion asymmetry and destabilizing moments, which can lead to loss of hover stability and maneuverability. This paper introduces HEXmorph, a novel fault-tolerant hexacopter design that employs geometric morphing through arm sweeping to redistribute thrust and counteract the effects of single and adjacent rotor failures. The proposed system integrates two optimization strategies: Moment Optimized Solver (MOS) and Center of Mass Optimized Solver (COS), tailored to minimize attitude changes and maintain stability during morphing. A feed-forward neural network is utilized to predict servo angles for arm morphing, ensuring real-time adaptability. The morphing mechanism is governed by a global event-triggered sliding mode control, which locks servo movements within a predefined error threshold. At the same time, system stability is guaranteed using a Modified Nonsingular Terminal Sliding Mode Controller (MNTSMC). Simulation and experimental results demonstrate the ability of HEXmorph to maintain near-zero attitude static hover and maneuverability, even under scenarios involving up to two adjacent rotor failures. By combining hardware adaptability with robust control strategies, HEXmorph significantly advances fault tolerance for multi-rotor aerial systems.