This paper deals with the problem of designing a\uddistributed fault detection and isolation methodology for nonlinear\uduncertain large-scale discrete-time dynamical systems. As a\uddivide et impera approach is used to overcome the scalability issues\udof a centralized implementation, the large scale system being\udmonitored is modelled as the interconnection of several subsystems.\udThe subsystems are allowed to overlap, thus sharing some\udstate components. For each subsystem, a Local Fault Diagnoser is\uddesigned, based on the measured local state of the subsystem as\udwell as the transmitted variables of neighboring states that define\udthe subsystem interconnections. The local diagnostic decision is\udmade on the basis of the knowledge of the local subsystem dynamic\udmodel and of an adaptive approximation of the interconnection\udwith neighboring subsystems. The use of a specially-designed\udconsensus-based estimator is proposed in order to improve the\uddetectability and isolability of faults affecting variables shared\udamong overlapping subsystems. Theoretical results are provided\udto characterize the detection and isolation capabilities of the proposed\uddistributed scheme. Finally, simulation results are reported\udshowing the effectiveness of the proposed methodology.
展开▼
机译:本文针对非线性不确定的大型离散动力系统设计\分布式分布式故障检测与隔离方法。由于采用了“ uddivide et impera”方法来克服集中式实现的可伸缩性问题,因此将要监控的大型系统建模为多个子系统的互连。ud允许子系统重叠,从而共享一些udstate组件。对于每个子系统,根据子系统测得的局部状态以及定义子系统互连的相邻状态的传输变量,设计局部故障诊断程序。基于局部子系统动态模型/ udmodel的知识以及与相邻子系统的互连\ ud的自适应近似,可以做出局部诊断决策。为了提高影响共享变量/重叠子系统的变量的错误检测能力和可隔离性,建议使用一种特殊设计的基于“共识”的估计器。提供理论结果以表征所提出的\ ud分布方案的检测和隔离能力。最后,报告的仿真结果\证明了所提出方法的有效性。
展开▼