Fault detection and identification method using observer-based residuals

Manufacturing machinery is becoming increasingly complicated, and machinery breakdowns not only reduce efficiency, but also pose safety hazards. Due to the needs for maintaining high reliability within facility operation, various methods for condition monitoring are suggested as the importance of maintenance has increased. Among the various prognostics and health management (PHM) techniques, this paper introduces a model-based fault detection and isolation (FDI) technique for the diagnosis of machine health conditions. The proposed approach identifies faults by extracting fault signal information such as the magnitude or shape of the fault based on a defined relationship between a fault signal and observer theory. To validate the proposed method, a numerical simulation is conducted to demonstrate its fault detection and identification capabilities in various situations. The proposed method and data-driven methods are then compared with regard to their fault diagnosis performance.