Analysis of charge-exchange spectroscopy data by combining genetic and Gauss–Newton algorithms. (November 2015)

• Record Type: Journal Article
• Authors: Qian, Ma; Haoyi, Zuo; Yanling, Wei; Liang, Liu; Wenjin, Chen; Xiaoxue, He; Shirong, Luo
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 166(2016:Apr.)
• Page Start: 74
• Page End: 80
• Publication Date: 2015-11
• Subjects: Active charge exchange spectroscopy -- Diagnostic of hot fusion plasmas -- GAGN; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2015.07.014
• Languages: English
• ISSNs: 0022-4073

Abstract

Abstract: The temperature and rotation velocity profile of ions in a tokamak are two characteristic parameters that reflect the plasma׳s behavior. Measurement of the two parameters relies on analyzing an active charge exchange spectroscopy diagnostic. However, a very challenging problem in such a diagnostic is the existence of interfering spectral lines, which can mislead the spectrum analysis process. This work proposes combining a genetic algorithm with the Gauss–Newton method (GAGN) to address this problem. Using this GAGN algorithm, we can effectively distinguish between the useful spectrum line and the interfering spectral lines within the spectroscopic output. The accuracy and stability of this algorithm are verified using both numerical simulation and actual measurements. Highlights: This paper presents an alternative computational approach to fit charge exchange spectra. The GN is embed into GA, forming a new algorithm GAGN. The GAGN algorithm can get the precise value without noise, but GA and GN cannot.