Limited Information Based Non-Additive Fault Detection and Diagnosis for CARIMA Model via Fuzzy Logic

Controlled auto-regressive integrated moving average (CARIMA) model based non-additive fault detection and diagnosis (FDD) for discrete system is proposed in this paper. A new estimate algorithm is developed to detect non-additive faults via limited information. It is designed based on previous limited inputs and outputs of system and used to obtain residual signals responding to the non-additive fault. Furthermore, the fuzzy logic based decision-making scheme and related criterion are proposed to make the on-line detection for non-additive fault of discrete system. The simulation results indicate that the new approach can detect non-additive faults of discrete system.