摘要
赋时事件图(TEG)是一类用于描述同步特性的重要的Petri网.本文提出并研究了基于赋时事件图的故障诊断问题.文中定义了两类系统故障类型:变迁失效(Invalidation)故障和变迁时间延迟(Time_lag)故障.基于事件图关联矩阵,通过研究变迁失效故障的可观测传播特性,引入了故障特征向量的概念.基于此,给出了失效故障的可诊断性的充分必要条件.对于时延故障,提出了一种基于变迁触发时间估计进行故障诊断的简单算法.并基于此算法,研究了时延故障的可诊断条件.
Timed event graphs （TEG） are an important subclass of Petri nets that are well adapted to model synchronizations. The problem of fault diagnosis for timed event graph is studied in this paper. Two different system faults are defined, transition invalidation fault and transition time-lag fault. The observable propagating path of invalidation fault is studied based on the relevance matrix of event graph, and fault character vector is firstly introduced to mark the propagating character of transition invalidation fault. Based on that, a sufficient and necessary condition of diagnosability for invalidation fault is given. For the time-lag fault, a simple fault diagnosis algorithm based on transition firing time estimation is given. Finally,the diagnosable condition is studied.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
2005年第4期609-614,共6页
Control Theory & Applications
基金
SupportedbyNationalScienceFoundationofChina(60074012)
theNationalFundamentalResearchFunds(G1998020310).
