Fault detection and classification using Kalman filter and genetic neuro-fuzzy systems

In this paper, an efficient scheme to detect the unprecedented changes in system reliability and find the failed component state by classifying the faults is proposed using kalman filter and hybrid neuro-fuzzy computing techniques. A fault is detected whenever the moving average of the Kalman filter residual exceeds a threshold value. The fault classification has been made effective by implementing a hybrid Genetic Adaptive Neuro-Fuzzy Inference System (GANFIS). By doing so, the critical information about the presence or absence of a fault is gained in the shortest possible time, with not only confirmation of the findings but also an accurate unfolding-in-time of the finer details of the fault, thus completing the overall fault diagnosis picture of the system under test. The proposed scheme is evaluated extensively on a two-tank process used in industry exemplified by a benchmarked laboratory scale coupledtank system.