Subspace based Fault Detection and Identification for LTI Systems

This paper presents a new Fault detection and Identification approach Connected to Subspace Identification (FICSI) for LTI systems with additive faults. FICSI parameterizes the influence of initial states on residuals by past I/O measurements, and hence avoids projecting a residual vector onto the left null space of the extended observability matrix. It is proved that FICSI produces residuals more sensitive to faults than the classic parity space approach (PSA) does. Another novelty of this contribution is the unbiased fault estimation algorithm developed in the framework of FICSI, which is not possible in PSA without restricting the nature of faults. The difference of FICSI from the existing fault detection and estimation approaches based on PSA or moving window estimation (MHE) can also be attributed to the fact that FICSI does not require a state space model, but a sequence of Markov parameters mapping the I/O measurements to the residual, which can be estimated in closed-loop.