Computer Engineering and Applications ›› 2016, Vol. 52 ›› Issue (21): 252-257.

Previous Articles     Next Articles

Multiple model fault diagnosis method for quadrotor UAV

XU Xuesong   

  1. School of Electrical Engineering & Automation, East China Jiaotong University, Nanchang 330013, China
  • Online:2016-11-01 Published:2016-11-17

四旋翼无人机的多模型故障诊断

徐雪松   

  1. 华东交通大学 电气与自动化工程学院,南昌 330013

Abstract: Quadrotor Unmanned Air Vehicle(UAV) is a close-coupled and strong nonlinear object. It is very important to find out faults precisely and regulate the controller quickly for maintaining safety of the quadrotor UAV. In this paper, an online fault diagnosis method based on multiple model algorithm is presented for actuator fault detection of the quadrotor UAV. In this method, a fault model set is built according to different fault modes. The operating mode is online monitored by parallel operation of the model set. Simulation results show with this method the parameters of fault model can be estimated quickly and precisely for partial loss of effectiveness faults of singular actuator.

Key words: quadrotor Unmanned Air Vehicle(UAV), fault diagnosis, multiple model filter

摘要: 四旋翼无人机是一个强耦合非线性被控对象,准确地判断故障情况,并及时调整控制器参数对无人机的安全飞行具有重要的意义。针对四旋翼无人机的执行器部分失效故障,提出了一种多模型在线故障诊断方法。该方法根据不同故障模式建立故障模型集,通过模型集的并联运行对无人机工作模式进行在线监测。仿真结果表明,对于单旋翼执行器部分失效故障,该方法能够快速准确地获得故障模型参数。

关键词: 四旋翼无人机, 故障诊断, 多模型滤波