Fault diagnosis for dynamical systems using soft computing

This paper describes fault diagnosis for small-scale dynamical systems using a new fast fuzzy neural network with general parameter (GP) based radial basis function network (RBFN) and a novel scheme of linear regression and time delay neural network. They are applied to fault diagnosis for an automobile transmission gears. Finally, a fault diagnosis scheme for large-scale and complex dynamical systems using fast fuzzy neural network with GP learning and immune networks is developed. It has been practically applied to sensor fault diagnosis for an uninterruptible power supply feedback control system. It is confirmed by simulations that the proposed methods are feasible for fault diagnosis