The Study of Inverse Heat Conduction Problems Using Input Estimation Method

• Main Authors: Chen, Tsung-Chien, 陳聰堅
• Other Authors: Tuan, Pan-Chio
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/41053034490119899513

博士 === 國防大學中正理工學院 === 國防科學研究所 === 92 === This work presents an inverse method based on the input estimation method including a finite-element method for solving the linear inverse heat conduction problem (IHCP). The input estimation method is comprised of the Kalman filtering technique and a recursive least-squares estimator. The state equation is derived using the finite-element method. Irregularly shaped boundaries can be approximated using elements with straight sides or matched exactly using elements with curved boundaries. Therefore, the proposed method is not limited to nice shapes with easily defined boundaries. The proposed method’s superiority is demonstrated using several cases involving a one-dimensional jet flow heat flux and heat flux on the barrel inner wall problems, two-dimensional plate, hollow cylinder and nozzle throat insert IHCP. The results show that this method has stability and good accuracy. The proposed method can solve more types of practical IHCP problems. This nonlinear IHCP study uses the Extended Kalman Filter (EKF) to replace simple Kalman filter. A definition of the Kalman tuning parameter is used to analyze the interactive relationship between the measurement noise and modeling error variance. The inverse method is helpful in obtaining satisfying estimation results. The simulation results demonstrate good performance and accuracy in tracking the unknown heat source in a thermal system.