ライブラリ登録: Guest

ISSN Online: 2377-424X

ISBN Print: 978-1-56700-474-8

ISBN Online: 978-1-56700-473-1

International Heat Transfer Conference 16
August, 10-15, 2018, Beijing, China

APPLICATION OF KF-RLSE METHOD FOR REAL-TIME RECONSTRUCTING THE TIME-VARYING HEAT FLUX IN ROSSELAND APPROXIMATION MODEL

Get access (open in a dialog) DOI: 10.1615/IHTC16.rti.023557
pages 8391-8403

要約

Real-time retrieval of the time-varying heat flux in the participating media is still a huge challenge, due to the intelligent algorithm and traditional gradient algorithm have a complex computation process. A numerical algorithm is developed to real-time and on-line reconstruct the high-magnitude and time-dependent heat flux on the surface of the participating media. The numerical method is consists of the Kalman filter technique and recursive least squares estimator (KF-RLSE). The KF has been applied to produce a regression model. Based on the regression model, the RLSE is employed to predict the time-varying heat flux on the surface of the media. Radiation energy can only transmit a short distance before it is attenuated, when the temperature gradient inside the medium is small and the optical thickness of the medium is large. In the case, the integral-differential form of radiation transfer equation can be converted to the thermal conductivity form of diffusion equation, because the movement of photons can be seen as a diffusion process, which is similar to molecular transport. Therefore, the Rosseland approximation is used to calculate the radiative heat flux in the energy equation. Finally, the effect of measurement noise covariance, process noise covariance, and thickness of the model on the accuracy and stability of the reconstruction results are discussed. The retrieval results showed that the accuracy of reconstruction results is improved as the process noise covariance increases.