Observer-Based Fault Diagnosis for Re-entrant Manufacturing Systems

This paper investigates the fault detection and diagnosis problem of re-entrant manufacturing systems (RMSs) in the presence of workstation faults. In order to visualize the dynamical behaviors of RMSs under the influence of impulse-type faults, a hyperbolic hybrid partial differential equation (PDE) continuum model is developed. Based on the diagnostic observer, a residual evaluation logic composed of an evaluation function and thresholds is proposed for fault detection. Upon detecting the faults, an adaptive observer which utilizes an adaptive fault estimation algorithm is proposed to accomplish the estimation of faults. By introducing a piecewise Lyapunov function candidate, the asymptotic stability (AS) conditions are presented to guarantee the boundness of the estimation error. The feasibility and applicability of the proposed strategy are validated through numerical simulations.