Comparisons and Improvements Concerning the Accuracy and Robustness of Inverse Heat Conduction Algorithms

摘要

A programming mistake that resulted in the misleadingly poor performance of the Sequential Estimation algorithm has been found and corrected. The authors apologize for any inconvenience this error may have caused. Although the conclusions in section 10 of the paper are still valid, it is shown below that the performance of the Sequential Estimation algorithm compares well with the SVD and Conjugate Gradient methods. 1) Page 189, second sentence four paragraph should read: 'Although Table 2 shows that the former obtained a better estimated heat flux than the latter, an even better estimated result could be obtained if the number of future time steps is increased.' 2) Results for Table 2 in page 193 are updated below by using the same pseudo-random noise vector for all simulations. The pseudo-random noise was added to the ideal, noiseless temperature data to simulate a noisy temperature measurement. The noise was generated using a Gaussian pseudo-random number generator in Matlab with an initial state value of 1.