A meshless FDM applied to a posteriori error analysis of experimental data by physically based global method approximation

摘要

In the paper proposed is extension of the recently developed and investigated new approach to physically based enhancement of experimental data by the so called global method (GM) to the field of a posteriori error estimation of experimental data. Unlike the classical methods of experimental data analysis, which frequently limit themselves to merely statistical estimates of the possible experimental error bounds, this new approach makes it possible to estimate the actual experimental error. It is possible thanks to application of the concept of physically based approximation for data enhancement (smoothing), which enriches the experimentally determined data with a priori knowledge coming from all the possible sources, including theory, statistics, other experiments performed in parallel, or even certain heuristic postulates should they be rendered reasonable. In the paper presented are basic concepts, a general formulation of the method followed by exemplary results for a test problem.