Charge Exchange Spectral Data Analysis Method Research On Tokamak

Charge-exchange recombination spectrum (CXRS) diagnosis is a very powerfulActive diagnostic tool for non-contact measurement of the internal parameters ofhigh-temperature plasma. The complexity and importance of the CXRS spectral signalhas been widely recognized by the fusion community. Simulation of CXRS composition,can be decomposed into standard non-linear curve (such as Gaussian distribution) whichcontains parameter informations of local plasma. The main content of this paper isbased on MATLAB and IDL programming, analyses and studies CXRS data fitting andfitting algorithms.Firstly,This article briefly described the related content about the fusion,andintroduced the course of development of magnetic confinement of china and containerscalled tokamak to achieve nuclear fusion. Secondly it introduced spectrum radiationcharacteristics of a high-temperature plasma and described the advantage ofcharge-exchange recombination spectrum (CXRS) diagnosis compared with otherdiagnostic methods, its principle,CXRS diagnostic system on the HL-2A tokamak andhow to get the raw spectral data in detail.And then it introduced programming methodof least squares and spectral data processing based on the MATLAB, genetic algorithmsprogramming and spectrum data processing based on the MATLAB, programmingmethod of least squares and spectral data processing based on the IDL. Finally analyzeand summarize the experimental algorithm and results. HL-2A of Southwest Institute ofNuclear Physics for CXRS experimental data were fitted to obtain the spatialdistribution of ion temperature and rotational speed of the different channels in the sameframe. This article is the first time to use the genetic algorithm include five Gaussianfitting curves,successes to apply the algorithm of least squares for CXRS dataprocessing to the IDL data processing platform and compare the results of the spectrumdata processing with the MATLAB-based least squares algorithm for CXRS which islong-term development. The results show that the fitting results based on MATLABgenetic algorithm and IDL least squares algorithm are basically the same. The fittingeffect by genetic algorithm is also very good. The method of least squares spectrum data processing in IDL environment is more efficient than in MATLAB environment.