基于局部保留投影的卫星姿控系统故障检测  

Fault Detection of Satellite Attitude Control System Based on Local Preserved Projection

作　　者：施常勇 胡立生[1] 郭祥 Shi Changyong;Hu Lisheng;Guo Xiang(School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China;Shanghai Institute of Spaceflight Control Technology,Shanghai 201109,China;Shanghai Key Laboratory of Aerospace Intelligent Control Technology,Shanghai 201109,China)

机构地区：[1]上海交通大学电子信息与电气工程学院,上海200240 [2]上海航天控制技术研究所,上海201109 [3]上海市空间智能控制技术重点实验室,上海201109

出　　处：《航天控制》2023年第2期53-59,共7页Aerospace Control

摘　　要：采用流形理论,提出了基于流形学习的局域保持投影对姿态系统运行数据进行特征学习,并利用统计特性进行故障检测的方法。首先,利用局域保持投影方法发现统计量的局部流形结构。该投影方法的本质是试图将原始空间中的闭合点映射到低维空间中的闭合点;其次,构建T2和平方预测误差(SPE),利用正常数据训练结果进行统计,并采用核密度估计(KDE)确定故障控制限,实现了对故障的检测。通过数据仿真验证所提方法,典型故障检测率为100%。Based on manifold theory,a method of feature learning is proposed,which is based on the local preserving projection of manifold learning to the operating data of the attitude system and fault detection based on the statistical characteristics.Firstly,the local manifold structure of the statistic is found by the local preserving projection method.The purpose of the projection method is used to map the closure points in the original space to the closure points in the low dimensional space.Secondly,T2 and square prediction error(SPE)are established,normal data training results are used for statistics,and kernel density estimation(KDE)is used to determine the fault control limit,and fault detection is realized.Through data simulation verification,the proposed method can effectively realize fault detection by a typical fault detection rate of 100%.

关 键 词：流形学习 局部保留投影(LPP) 核局部保留投影(KLPP) 姿态控制系统 故障检测 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计]