Inverse heat conduction problem: Sensitivity coefficient insights, filter coefficients, and intrinsic verification

摘要: Abstract The inverse heat conduction problem is the estimation of time and/or space dependence surface flux or temperature utilizing interior measurements at discrete times locations. This ill-posed since it very sensitive to omnipresent measurement errors. Many solution methods have been proposed including exact-matching, function specification, Tikhonov regularization, iterative conjugate gradient, steepest descent and singular value decomposition. In this paper, tools provided by scaled sensitivity coefficients, digital filter intrinsic verification are used investigate compare several these methods. utility filters designed for on-line instrumentation “continuous” in manufacturing settings also demonstrated.