Fault Diagnosis and Reconfigurable Control of a Pressurizer in a Nuclear Power Plant

Abstract

In this paper, an integrated approach to the design of fault diagnosis and reconfigurable fault-tolerant control system against actuator and sensor faults in a nuclear power plant pressurizer is proposed. The scheme uses a reconfigurable proportional-integral (PI) control structure and a setpoint adjustment technique to achieve required performance at both the dynamic and steady-state conditions. A singular value decomposition (SVD) based eigenstructure assignment (EA) technique is used to achieve on-line automatic redesign of the controller. Fault diagnosis and controller reconfiguration mechanisms are carried out using statistical hypothesis tests based on the information from a two stage adaptive Kalman filter. The proposed approach has been evaluated under different type of Unanticipated actuator faults: e.g. abrupt and incipient faults; total and partial faults; single, multiple and consecutive faults under both set-point regulation and set-point tracking. Through set-point tracking, specified performance degradation in the presence of actuator fault. can be eexplicitly considered. Detection and isolation of sensor faults are achieved by voting among outputs from several hardware redundant sensors. The effectiveness of the approach is demonstrated via simulation using a pressurizer model.