| Summary: | 碩士 === 國立中興大學 === 應用數學系所 === 100 === This study presents a hybrid numerical method to inverse estimation of the thermal conductivity in a one-dimensional long circular tube. The hybrid algorithm is based on the Savitzky-Golay digital filter, to smoothen the measurement error data, and the second order finite difference method, to calculate inversely the thermal conductivity. In addition, a polynomial regression used to obtain the empirical formula for the thermal conductivity.
The influences of measurement errors w, number of the measurement points j, and breadth of the smoothed window N on predicting the thermal conductivity, are discussed in detail. The results show that the accuracy of the estimated value is increased as the number of measurement point j or the breadth of the smoothed window N increases.The applied Savitzky-Golay digital filter can significantly reduce the error of the predicted value for various measurement errors w. It is thus evident that the hybrid numerical algorithm is a simple and effective method to predict the thermal conductivity.
|
|---|
