基于深度学习的无人机故障诊断方法研究  

Research on Fault Diagnosis Method of UAV Based on Deep Learning

作　　者：李炜[1,2,3] 崔佳佳[1,2,3] LI Wei;CUI Jiajia(School of Electrical Engineering and Information Engineering,Lanzhou University of Technology,Lanzhou 730050;Gansu Provincial Key Laboratory of Advanced Process Control,Lanzhou 730050;National Demonstration Center for Experimental Electrical and Control Engineering Education,Lanzhou University of Technology,Lanzhou 730050)

机构地区：[1]兰州理工大学电气工程与信息工程学院,兰州730050 [2]甘肃省工业过程先进控制重点实验室,兰州730050 [3]兰州理工大学电气与控制工程国家级实验教学示范中心,兰州730050

出　　处：《计算机与数字工程》2019年第5期1249-1253,共5页Computer & Digital Engineering

基　　金：国家自然科学基金项目(编号:61364011)资助

摘　　要：由于四旋翼无人机(quadrotor UAV)系统耦合性较高,难以建立准确的系统模型,致使故障诊断准确率较低,同时由于无人机工作环境复杂,易受外界和内部噪声影响,使得大多数方法存在鲁棒性不足的问题。所以,论文提出栈式降噪自编码(Stacked Denoising Autoencoder,SDA)的故障诊断方法,不仅不过分依赖系统模型,而且增强其方法的鲁棒性,提升故障诊断的准确率。仿真结果表明,基于SDA的无人机故障诊断方法,能准确可靠地判断执行器的故障类型,并有效提升了四旋翼无人机的安全可靠性。Due to the high coupling of the quadrotor UAV system,it is difficult to establish the appropriate system model,the accuracy of fault diagnosis is low. At the same time,due to the complex working environment of UAV,it is easy to be affected by the external and internal noise,which makes most methods have the problem of insufficient robustness. Therefore,this paper proposes the Stacked Denoising Autoencoder(SDA)fault diagnosis method. It not only does not rely too much on the system model,but also enhances its robustness and improves the accuracy of fault diagnosis. The simulation results show that the UAV fault diagnosis meth. od based on SDA can correctly and reliably judge the fault type of the actuator,and effectively improve the safety and reliability of the four rotor UAV system.

关 键 词：四旋翼无人机 栈式降噪自编码 鲁棒性 

分 类 号：V279[航空宇航科学与技术—飞行器设计]