Thermal condition monitoring in a chemical looping combustion reactor for real-time operation diagnosis

Abstract

With the increasing thermal input of chemical looping combustion (CLC) reactor, the monitoring and diagnosis of its internal boundary conditions become important issues for operation safety. Based on the discrete heat transfer model of reactor wall, Kalman filter and fuzzy inference were combined as fuzzy inference-based augmented Kalman filter (FI-AKF) and fuzzy inference-based Kalman filter coupled with weighted recursive least squares algorithm (FI-KFW) for the real-time monitoring of CLC reactor. Simulations were carried out to validate the feasibility of these two methods. Under both normal and abnormal conditions, the FI-KFW could exhibited satisfying performances for the internal heat flux monitoring. Number of the measurement points and intensity of measurement noises were changed numerically to investigate their effects on the monitoring results of FI-KFW. Results demonstrated that FI-KFW had strong ability to resist the ill-posedness of the monitoring process and it was capable for the real-time monitoring of the chemical looping combustion reactor, which could offer reliable information for operation diagnosis.