In this work a comparative study of different time domain fitting algorithms in multiexponential responses is presented, particularly in the case of thermally activated processes. We have chosen three algorithms found in the literature and another one that we propose. The study has been carried out by computer simulation with the same conditions of signal-to-noise ratio and response complexity. To present the results we have tested each algorithm with a signal composed of two thermally activated processes and a noise superposed.
KronmüllerH., Nachwirkung in Ferromagnetika, Springer-Verlag, 1968.
2.
LangD.V., J. Appl Phys.45 (1974), 3014–3023.
3.
FioraniD., TestaA.M. and TejadaJ., Flux motion in Bi- and Ti-based superconductors, Appl. Supercond. 1 (1993), 935–945.
4.
MironovS.L., A simple non-iterative procedure for fitting multiexponential functions, J. Neurosci. Methods38 (1991), 243–246.
5.
MartinJ.L., MaconochieD.J. and KnightD.E., A novel use of differential equations to fit exponential functions to experimental data, J. Neurosci. Methods51 (1994), 135–146.
6.
AbramowitzM. and StegunI.A., Handbook of Mathematical Functions, Dover Publications, New York, 1972.
7.
FondadoA. and RivasJ., Application of an algorithm for fitting exponential functions to magnetic relaxation phenomena, IEEE Trans. Magn. 29(6) (1993), 3010–3012.
8.
AlejosO., de FranciscoC.HernándezP. and MuñozJ.M. (accepted for publication).
9.
YeramianE. and ClaverieP., Analysis of multiexponential functions without an hypotesis as to the number of components, Nature326 (1987), 169–174.
10.
KukulinV.I., KrasnopolskyV.M. and HorácekJ., Theory of Resonances. Principles and Applications, Kluwer, 1989.
11.
OppenheimA.V., WillskyA.S. and YoungI.T., Signals and Systems, Prentice Hall, 1983.
