强噪声下基于混沌技术的高超声速飞行器传感器故障诊断  

Fault Diagnosis Based on Chaos System under Strong Noise

作　　者：黄鹏飞 屈剑锋[1] 柴毅[1] 陈小龙 刘切 HUANG Pengfei;QU Jianfeng;CHAI Yi;CHEN Xiaolong;LIU Qie(School of Automation,Chongqing University,Chongqing 400044,China)

机构地区：[1]重庆大学自动化学院,重庆400044

出　　处：《宇航学报》2023年第8期1203-1212,共10页Journal of Astronautics

基　　金：国家重点研发计划(2021YFB1715000)。

摘　　要：针对高超声速飞行器强非线性和强噪声问题,开展了迎角传感器故障诊断方法研究。基于混沌系统的噪声免疫性质,通过对分数阶混沌系统参数和阶次的设计,提出了一种故障信号特征提取方法,该方法可有效降低强噪声对故障特征提取和故障诊断带来的不利影响;采用结合系统模型和机器学习的高超声速飞行器传感器故障诊断方法,构建了高超声速飞行器迎角传感器故障诊断分类器。仿真结果验证了所提出的分数阶混沌系统在故障特征信号提取方面的有效性,可显著提升故障诊断的准确率。A study on fault diagnosis methods for angle of attack sensors is made to address the strong nonlinearity and noise issues of hypersonic aircraft.Based on the noise immunity properties of the chaos system,a fault signal feature extraction method is proposed by designing the parameters and order of the fractional-order chaos system.This method can effectively reduce the adverse effects of strong noise on fault feature extraction and fault diagnosis.A hypersonic aircraft angle of attack sensor fault diagnosis classifier is constructed using a combination of system models and machine learning for hypersonic aircraft sensor fault diagnosis.The simulation results validate the effectiveness of the proposed fractional-order chaos system in extracting fault signal features,which can significantly improve the accuracy of fault diagnosis.

关 键 词：高超声速飞行器 传感器故障 故障诊断 分数阶系统 混沌系统 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]