摘要
本文提出一种基于相空间重构和故障跟踪算法的复杂电子系统故障诊断方法,该方法将复杂电子系统中参数变异演化造成的不稳定因素作为跟踪对象,并将其视为隐藏在“快变”系统状态矢量中的“慢变”状态矢量,利用相空间重构和局部线性化模型以及无故障系统的观测数据构造用于故障跟踪的参考模型,用数据替代法生成系统故障数据,借助于对系统状态矢量在相空间演化过程中相对参考模型误差的跟踪与估计,实现系统故障的检测与识别。混沌时间序列的相空间重构采用了复自相关和Γ-test 的嵌入维、时间延迟联合算法,所得到的结果为准最佳的嵌入维、时间延迟。实验结果表明,该故障诊断方法可准确地检测出复杂电子系统中的故障现象。
A fault-diagnosis method based on the phase space reconstruction and fault tracking approaches for the complex electronic system is proposed. This method treats the unstable factors arising from the variation of parameters in complex systems as the tracking objects, which is considered as the “slow dynamic” feature vector hidden in the “fast dynamic” system feature vectors. A reference model for the fault tracking is set up based on the Phase Space Reconstruction and Local Linear Model, as well as the testing data of a faultless system. The data of system fault are generated using the data surrogate algorithm. The system faults are detected and identified by tracking and estimating the errors, which are caused by the system feature vectors evolved in the phase space comparatively to the reference model. The approach for the phase space reconstruction of chaotic time series is a combination algorithm of multiple autocorrelation and Γ -test, by which the quasi-optimal embedding dimension and time delay can be obtained. The experimental results depict that this fault diagnosis method can correctly detect the fault phenomena of electronic system.
出处
《电路与系统学报》
CSCD
2004年第5期76-80,共5页
Journal of Circuits and Systems
关键词
复杂电子系统
故障诊断
混沌时间序列
相空间重构
complex electronic system, fault diagnosis, chaotic time series, phase space reconstruction
