摘要
当需要诊断的故障数目较多、扩展故障参数后状态空间不可观测时 ,用一个故障诊断模型无法实现多故障诊断。为此 ,在分析了基于故障参数与状态联合估计的故障诊断方法原理的基础上 ,根据多个故障诊断模型之间存在的内在关系 ,提出了一种新的多故障检测和诊断方法 ,认为在某一段时间内 ,如有 n- 1(n为总故障数目 )个不同的两故障参数模型发出同一故障报警 ,则该故障报警有效。该故障诊断方法综合多个故障诊断模型的诊断结果 ,从而使一类复杂系统的多故障诊断成为可能。理论分析和仿真实验的结果证明该文所提出的方法是可行的 ,并且在工程上是易于实现的。
A novel multi fault diagnosis algorithm is presented based on the relationship between different fault diagnosis models using joint estimation of states and fault parameters. This method can detect multiple faults which can not be detected by one diagnosis model because of the extended state space due to all the fault parameters. The method synthesizes all the two fault parameter models and considers the same fault alarm given by n -1( n is the number of faults to be detected) two fault parameter models in a period of time for the valid fault alarm. The results are validated by a theoretical analysis and a simulation.
出处
《清华大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2001年第12期92-94,共3页
Journal of Tsinghua University(Science and Technology)