Improved fault detection and identification for nonlinear hybrid systems using self switched CDKF

The main objective of this paper is to develop a novel and improved method of fault detection for a class of nonlinear hybrid plants with emphasis on identification. An approach has been presented here to evolve techniques which enable efficient fault detection and their type and range identification using a bank of self-switched Central Difference Kalman Filter which is numerically simpler than EKF. Validation of developed methods would be done with hybrid models of real-life systems. Different ranges of clogging faults and leakage faults of different radii of a three tank system have been considered here to demonstrate the effectiveness of the scheme. The inflow of tank 1 has been considered noisy. Simulation results show that different ranges of clogging faults and leakage faults can be successfully detected and their ranges can be identified using this scheme. It has been shown that the CDKF based scheme not only can detect the fault accurately but also identify its type and range properly.