Solution of inverse radiation-conduction problems using a Kalman filter coupled with the recursive least-square estimator

This paper presents a non-intrusive inverse heat transfer procedure for predicting the time-varying heat flux on the surface of semitransparent media. Inverse radiation-conduction problems were analyzed using a Kalman filter coupled with a recursive least-square estimator (KF-RLSE). Since its unique capability to make fast predictions, KF-RLSE can be easily integrated to existing real-time control systems of industrial facilities. The performance of KF-RLSE was examined thoroughly in a series of numerical simulations in two semitransparent materials (i.e. the glass with black coating and the ceramic of zirconium dioxide ZrO2) to extract the time-varying surface heat flux on-line from the measured temperature history at boundary. Results showed that the proposed method can predict the unknown boundary flux with an acceptable error. The influence of different parameters on the accuracy and stability of the predicted heat flux was also investigated. Results indicated that the sensor location, process noise covariance and absorption coefficient exerted stronger effects on retrieval results compared with other parameters.