基于切换与随机混合模型的网络化奇异系统故障检测

项目名称： 基于切换与随机混合模型的网络化奇异系统故障检测

项目编号： No.61203085

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 自动化学科

项目作者： 万雄波

作者单位： 中国地质大学（武汉）

项目金额： 25万元

中文摘要： 网络化系统由于其诸多优点而在各领域得到了广泛应用。同时，由于对象规模和复杂度增加，系统发生故障则可能带来严重的危害。因此，网络化系统故障检测受到了人们越来越多重视。本项目拟提出若干切换与随机混合模型同时描述多种网络现象。进而引入一些性能指标,将网络化系统故障检测问题转化为最优化问题。利用切换Lyapunov函数法或平均驻留时间法，我们将给出滤波误差系统渐近均方稳定且满足给定性能指标约束的充分条件，并且当上述条件有可行解时清晰地给出残差产生器参数。考虑到奇异系统比正常系统更具一般性且在许多情形下更适合描述物理对象，我们将利用严格系统等价变换等方法将上述基于标准状态空间系统的故障检测结论推广到奇异系统情形。本研究旨在探索网络化系统建模新方法，建立故障检测新结论，分析网络诱导因素和系统奇异性对故障检测耗时量、误报率与漏检率的影响。

中文关键词： 复杂网络系统；故障检测；网络化系统；基因调控网络；基因振子网络

英文摘要： Networked systems have been finding application in a broad range of areas due to their many advantages. Meanwhile, with the increasing scale and complexity of the plants, the faults occurred may cause serious damage to the systems. Therefore, fault detection of networked systems has received more and more attention. This project will propose several switched and stochastic hybrid models to simultaneously describe multiple network-induced phenomena. Then, by introducing certain performance indexes, the fault detection problem of networked systems will be converted into an optimization problem. By the switched Lyapunov function approach or the average dwell time approach, sufficient conditions will be given for the filtering error system to be asymptotically mean-square stable with the prescribed performance indexs constraints. When the above conditions are feasible, the parameters of the residual generator will be explicitly given. Considering that the singular systems are more general and in many cases more proper to describe physical plants, the fault detection results based on the standard state-space systems will be extended to networked singular systems by strict system equivalence transformation approach, etc. The aim of this research is to explore new modeling approaches of networked systems, establish new

英文关键词： Complex networked systems；Fault detection；Networked systems；Genetic regulatory networks；Genetic Oscillator networks