基于系统仿真的故障检测与辨识技术研究

Research on Approach of Fault Detection and Identification Based on System Simulation

本文将系统仿真技术应用于故障检测与诊断(FDD),提出了通过实测信息在线交互减小解决理论轨线偏差的思想,建立了基于系统仿真FDD逻辑结构,针对MIMO系统和线性动态-测量系统构造了具有良好统计性质的故障辨识算法,为故障检测、故障辨识、故障时间的确定和故障模式识别提供了实现的可能性。本文的思想和方法不但可用于CVDS的故障检测与诊断,对大规模复杂系统的故障分析也具有借鉴和参考价值。

Based on system simulation,a new approach on fault detection and identification (FDD is suggested in this paper. The logical structure of FDI is built on system simulation process,in which the bias of theoretical trajectory is cut down by interactive mutual information. Under the multi-input multi-output system and dynamic-measurement system, a series of identification algorithms are given,which can be used to solve the following problem, such as fault detection,fault identification,estimation of fault time and its extent,etc. What is more,the idea given in this paper can be popularized to CVDS and LSS.