Decentralized Diagnosis via Structural Analysis and Integer Programming

Abstract : Centralized fault diagnosis architectures are sometimes prohibitive for large-scale interconnected systems such as distribution systems, telecommunication networks, water distribution networks, fluid power systems. This paper presents a decentralized fault diagnosis method for continuous systems that only requires the knowledge of local models and limited knowledge of their neighboring subsystems. The notion of Fault-Driven Minimal Structurally Overdetermined (FMSO) set is used as the corner stone of the design of residual generators for the design of decentralized fault diagnosis for systems that have constraints of confidentiality, distance or limited access to some information. Binary integer linear programming (BILP) is used to optimize the choice of FMSO sets in each local subsystem.