| Fault detection and isolation (FDI) has been an important field of research in the control engineering community for the last two decades. Its significance is based on enhancements in terms of safety, reliability and operating costs of the plant.;The efficacy and robustness of this technique is demonstrated by applying this FDI scheme to a jacketed continuously stirred tank reactor (JCSTR).;This research focuses on solving the failure detection and isolation (FDI) problem by developing a model-based approach using a parity equation implementation of directional residuals. This new approach is an extension of the generalized parity vector (GPV) technique based on the stable factorization. The present research has improved the approach in Viswanadham, Taylor and Luce [21] in three important aspects. First, a novel transformation matrix computation is presented that enhances the isolation properties of the FDI algorithm, i.e., increases the maximum number of faults that can be isolated and the number of disturbances that can be decoupled above the number of outputs of the system [7]. Second, disturbance decoupling is implemented in the stable factorization framework to make the residuals immune to measurable disturbance effects. Third, an adaptive threshold logic was developed and implemented to take into account modeling errors. |
