Simulated annealing and Bayesian inference applied to experimental stopping force determination

摘要

A data analysis method has been developed which allows to extract stopping force curves (a.k.a. stopping powers) from standard experimental backscattering spectra. The method is based on the simultaneous fit of a set of spectra, varying a parameterized stopping force curve. An initial fitting is achieved by means of a simulated annealing algorithm and then the method relies on a Bayesian inference algorithm that allows to obtain the uncertainties associated with the resulting stopping force curves. In order to avoid difficulties with the simulation of the experiments, a data reduction is performed on the spectra which makes simulation nearly trivial at the cost of a minor loss of information. This method, which has been implemented in an open source code named Hotstop, can be applied to light and heavy ions (pulse-height defect corrections and non-linear calibration are used), as well as to mono-elemental or compound targets.