Multi-Fault Dataset for Multirotor UAVs with Single- and Double-Magnitude Faults 核心 · 已核验

mxds0xme9xwzb154

"Reliable fault diagnosis is crucial for the safe operation of multirotor unmanned aerial vehicles (UAVs), yet publicly available datasets that include multiple simultaneous faults and explicit fault severity information remain limited. This paper introduces the Multi-Fault Dataset for Multirotor UAVs with Single- and Double-Magnitude Faults, a labeled time-series dataset designed to support the development and benchmarking of data-driven fault detection, isolation, and severity estimation algorithms. The dataset comprises 114,230 time-stamped samples recorded at 100 Hz from a nonlinear 6-DOF multirotor simulation, whose dynamics and sensor characteristics are matched to those of a Pixhawk Cube\u2013based autopilot. Each sample contains body angular rates (p, q, r), linear accelerations (ax, ay, az), and Euler angles (\u03d5, \u03b8, \u03c8), together with four normalized fault-magnitude indicators (m\u2081\u2013m\u2084) associated with each rotor. Fault magnitudes take discrete values in the range 0.05\u20130.40, in steps of 0.05, representing different severity levels. The data encompass 228 operating scenarios, including nominal operation as well as single-rotor and double-rotor faults, resulting in 57,547 nominal, 6,015 single-fault, and 50,668 double-fault samples. In addition, each rotor fault is annotated with a binary fault flag that explicitly marks the presence of a fault and can be used to structure sequential detection\u2013then\u2013estimation pipelines while avoiding unnecessary computation during nominal operation. Overall, the released file contains 11 time-series variables organized in a simple tabular format to facilitate direct use with classical machine learning and deep learning methods. By combining nominal, single-fault, and multi-fault conditions with multiple severity levels, this dataset provides a comprehensive benchmark resource for research on UAV health monitoring, fault-tolerant control, and prognostics."

本卡描述如含来自官方页的原文片段,其版权归原作者,不在本站 CC-BY 4.0 许可范围(见关于与许可)。

落地页

https://ieee-dataport.org/documents/multi-fault-dataset-multirotor-uavs-single-and-double-magnitude-faults

国内可访问性

国内直连:可达 (2026-07-11 检测) 非直连:可达 (2026-07-11 检测)
「直连超时」表示检测窗口内未完成,系慢或不稳定证据,不构成封锁证据。

发布年份

2025

发布方

IEEE DataPort

设备类型

rotorcraft_uav

PHM 任务

fault_diagnosis fault_severity_estimation

故障工况