Chapter 16 - Optimal state and parameter estimation for fault detection and diagnosis of a high-dimensional fluid catalytic cracking unit

Abstract

Optimal state and parameter estimation-based fault detection and diagnosis via quantitative model-based techniques are gaining considerable importance in plant operations. Efficient diagnosis of process plants can lead to considerable incentives in terms of high performance, reliability, and safety. Model-based fault diagnosis via state and parameter estimation relies on the principle that possible faults in the monitored process can be associated with specific parameters and states of the mathematical model of the process. Designing a state estimator for fault diagnosis of a high-dimensional, nonlinear, and strongly interacting fluid catalytic cracking unit (FCCU) with several subsystems like feed and preheat systems, cracking riser, reactor, regenerator, air blowers, and catalyst transport system is a challenging task. In this chapter, stochastic model-based estimators, namely, nonlinear discrete Kalman filter and unscented Kalman filter (UKF) are designed and applied for state estimation and fault parameter identification of the complex FCCU plant. The results analyzed for different fault scenarios exhibit the usefulness of stochastic quantitative model-based estimators, especially, the method of UKF for state estimation and fault diagnosis of FCCU.