摘要
为解决由于无人机故障数据缺乏且无故障标签而造成的无人机故障诊断、健康管理等工作开展困难的难题,提出了一种基于飞行控制系统仿真模型的全数字仿真的故障数据生成方法。详细分析了执行机构和传感器的故障模式及对应的数学模型。在明尼苏达大学无人机模型上,进行故障注入与故障仿真,生成故障数据,同时分析了不同故障类型、注入节点和幅值大小对故障数据生成的影响,总结了一套基于仿真模型生成故障数据的方法。所提出的方法利用了数字仿真操作简单,故障注入方式灵活、无人机机理表现充分等优点,可方便模拟无人机实际飞行过程中故障发生时间及故障模式的随机性,对改善目前无人机故障诊断领域缺少有效故障数据的现状具有重要意义。
In order to solve the difficulty of UAV fault diagnosis and health management caused by the lack of fault data and fault label,this paper proposes a full digital simulation fault data generation method based on the simulation model of flight control system.In the paper,the fault modes and corresponding mathematical models of actuators and sensors are analyzed in detail.Fault injection and fault simulation on the University of Minnesota UAV model are carried out to generate fault data.Meanwhile,the impacts of different fault types,injection nodes and amplitudes on fault data generation are analyzed.A set of fault data generation methods based on simulation model is summarized.The proposed method utilizes the advantages of simple digital simulation operation,flexible fault injection mode and full demonstration of UAV mechanism.The proposed method can easily simulate the randomness of the fault occurrence time and failure mode during the UAV actual flight process,which has great significance to improve the present situation of lacking effective fault data in UAV fault diagnosis field.
作者
彭宇
史书慧
郭凯
刘大同
Peng Yu;Shi Shuhui;Guo Kai;Liu Datong(School of Electronics and Information Engineering,Harbin Institute of Technology,Harbin150080,China;Automatic Test and Control Institute,Harbin Institute of Technology,Harbin150080,China)
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2019年第7期13-21,共9页
Chinese Journal of Scientific Instrument
基金
国家自然科学基金(61803121,61701131,61571160)资助项目
关键词
无人机
飞控系统
故障仿真
故障数据生成
UAV
flight control system
fault simulation
fault data generation
