Fault Detection, Isolation and Accommodation Using the Generalized Parity Vector Technique

This paper extends the generalized parity vector approach for fault detection and isolation presented in Omana and Taylor [2], [3], [4], to achieve sensor accommodation. In this study, this fault detection, isolation and accommodation technique is applied to a two-phase separator followed by a three-phase gravity separator model used in oil production facilities. This model simulates a large scale process, which allows the technique to be tested on a high dimensional space with more complex system dynamics. The fault management strategy is significantly improved by implementing a fault-size estimation and classification technique using the gpv magnitude signature. This fault characterization is refined by incorporating a recursive fault size recalculation algorithm based on the sensor accommodation error. Two different methods for sensor accommodation and fault size recalculation are proposed to take into account the software and hardware configuration in the plant.