ALFA
收藏资源简介:
ALFA数据集由卡内基梅隆大学机器人学院创建,专注于固定翼无人机的控制表面故障和异常检测。数据集包含47次自主飞行中的故障场景,涵盖了发动机、方向舵、副翼和升降舵等多种故障类型,总飞行时间为66分钟正常飞行和13分钟故障后飞行。数据集的创建涉及对飞行器的硬件和软件进行定制修改,以模拟各种故障情况。该数据集主要用于无人机故障检测和隔离(FDI)以及异常检测(AD)研究,旨在提高自主飞行操作的安全性。
The ALFA dataset was developed by the Robotics Institute at Carnegie Mellon University, focusing on control surface fault and anomaly detection for fixed-wing unmanned aerial vehicles (UAVs). It contains fault scenarios from 47 autonomous flights, covering multiple fault types such as engine failures, rudders, ailerons and elevators. The total flight duration consists of 66 minutes of normal flight and 13 minutes of post-fault flight. The creation of this dataset involved customized modifications to the aircraft's hardware and software to simulate various fault conditions. This dataset is primarily used for research on unmanned aerial vehicle fault detection and isolation (FDI) as well as anomaly detection (AD), aiming to improve the safety of autonomous flight operations.




