OpenDeckSMR涡轮风扇健康监测数据集
收藏资源简介:
该数据集由法国SystemX技术研究所联合赛峰科技开发,聚焦航空发动机健康监测领域,包含500条涡轮风扇发动机退化轨迹数据,每条轨迹最多含2000个时间步的10维健康指标和7维传感器测量值。数据通过OpenDeckSMR热力学模拟器生成,创新性地引入了三种退化速率、随机维护事件及四种飞行工况(巡航/起飞/爬升),并添加了符合真实场景的边界噪声。该数据集旨在解决稀疏传感器条件下发动机部件健康状态反演的病态逆问题,为数据驱动模型、贝叶斯滤波及自监督学习算法提供基准测试平台。
This dataset was jointly developed by the French SystemX Institute and Safran Technology, focusing on the field of aero-engine health monitoring. It contains 500 turbofan engine degradation trajectory datasets, where each trajectory includes up to 2000 time steps of 10-dimensional health indicators and 7-dimensional sensor measurements. The data was generated using the OpenDeckSMR thermodynamic simulator, and innovatively introduces three degradation rates, random maintenance events, four flight conditions (cruise, takeoff, climb) as well as boundary noise that conforms to real-world scenarios. This dataset aims to solve the ill-posed inverse problem of aero-engine component health state inversion under sparse sensor conditions, providing a benchmark test platform for data-driven models, Bayesian filtering and self-supervised learning algorithms.




