Book/Report

FZJ-2018-01903

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Multiexponentialanalyse von Messdaten durch eine automatische Abschälmethode und anschliessender nicht-linearer Least-Squares-Anpassung : Beschreibung des Fortran-Unterprogrammes Search und anderer Hilfsroutinen

Schröck-Pauli, C.FZJ* ; Schätzler, L.FZJ*

1982
Kernforschungsanlage Jülich, Verlag Jülich

Please use a persistent id in citations: http://hdl.handle.net/2128/17694

Report No.: Juel-1798

Abstract: This report is concerned with $\underline{multi-exponential-fitting}$ of a $\underline{model}$ $\underline{function}$ f(t) = $\sum^{n}_{j=1} a_{j} e^{- \alpha_{j}t}$ + $\eta$(t) (*) (a$_{j}$, $\alpha_{j}$>o, 1$\le$j$\le$n, $\eta$(t) = a+bt) to given experimental terms contained in (*) is $\underline{not}$ $\underline{known}$ $\underline{in}$ $\underline{advance}$. An automatic version of the well-known manually performed peeling technique is realized and implemented in the subroutine SEARCH. This program yields the above mentioned number n and initial values for the parameters a,b,a$_{j}$,$\alpha_{j}$,1$\le$j$\le$n, in addition, which serve as input data for a final non-linear fitting of model (*) by a convenient non-linear fit program, e.g. VARPRO (from FORTLIB of KFA) or VA13AD (from Harwell Subroutine Library). Moreover, auxiliary programs for evaluation of f, the partial exponential terms in f, and the appertaining possibly weighted Least squares functional F, respectively, as well as subroutines for determination of the first and second partial derivatives of f and F with respect to the parameters are made accessible. Characteristic examples of multi-exponential fitting to simulated and experimental data demonstrate the efficiency of the presented method.

---

Contributing Institute(s):

1. Publikationen vor 2000 (PRE-2000)

Research Program(s):

1. 899 - ohne Topic (POF3-899) (POF3-899)

---

Database coverage:

---