Sensor Fault Diagnosis Based on Improved Dynamic Structured Residual Approach in Dynamic Processes

A new sensor faults diagnosis method based on improved dynamic structured residual approach with maximized sensitivity (DSRAMS) is proposed for dynamic processes monitoring in this paper. The dynamic principal component analysis (DPCA) method is employed for system identification and model reduction. Extended incidence matrix is proposed to diagnosis the dynamical systems where one sensor fault will affect multiple elements of the measurement vector. Sensor faults sensitivity and critical sensitivity are defined, based on which an incidence matrix optimization algorithm is proposed. Simulation results in a dynamic process show the effectiveness of the proposed method.