VTOL Failure Detection and Recovery by Utilizing Redundancy

TL;DR

This paper presents a real-time, aircraft-independent fault detection system and an optimization-based control allocation method for hybrid VTOLs, enhancing their safety and fault tolerance through redundancy utilization and experimental validation.

Contribution

It introduces a novel fault detection system and a control recovery approach specifically designed for hybrid VTOL aircraft, addressing a gap in fault tolerance research.

Findings

Effective real-time fault detection demonstrated in simulations and experiments.

Control allocation successfully recovers from various actuator failures.

Redundancy utilization enhances hybrid VTOL safety and reliability.

Abstract

Offering vertical take-off and landing (VTOL) capabilities and the ability to travel great distances are crucial for Urban Air Mobility (UAM) vehicles. These capabilities make hybrid VTOLs the clear front-runners among UAM platforms. On the other hand, concerns regarding the safety and reliability of autonomous aircraft have grown in response to the recent growth in aerial vehicle usage. As a result, monitoring the aircraft status to report any failures and recovering to prevent the loss of control when a failure happens are becoming increasingly important. Hybrid VTOLs can withstand some degree of actuator failure due to their intrinsic redundancy. Their aerodynamic performance, design, modeling, and control have all been addressed in the previous studies. However, research on their potential fault tolerance is still a less investigated field. In this workshop, we will present a summary of our work on aircraft fault detection and the recovery of our hybrid VTOL. First, we will go over our real-time aircraft-independent system for detecting actuator failures and abnormal behaviors. Then, in the context of our custom tiltrotor VTOL aircraft design, we talk about our optimization-based control allocation system, which utilizes the vehicle's configuration redundancy to recover from different actuation failures. Finally, we explore the ideas of how these parts can work together to provide a fail-safe system. We present our simulation and real-life experiments.