|
PresentationsApproximation of absorption spectrum by Gaussian decomposition optimized with differential evolutionProkhorov General Physics Institute of the Russian Academy of Sciences, 119991, Moscow, Russia, Vavilova street 38, shkirinajuliana@gmail.com 1Lomonosov Moscow State University, 119991, Moscow, Russia, Leninskie Gory, 1 2Moscow Institute of Physics and Technology (National Research University), 141701, Dolgoprudny, Russia, Institutskiy per., 9 Gaussian functions are key to describing many physical processes and phenomena. In particular, they are used in modeling the absorption spectra of organic and inorganic molecules. In the simplest case, the spectrum can be approximated by a single Gaussian function defined by only three parameters. However, in most cases the absorption spectrum is described by a more complex line shape, and its approximation requires the use of a larger number of Gaussian functions. The procedure for selecting parameters in this case requires significant costs, resulting in the need to use evolutionary algorithms, such as differential evolution. We have developed software that allows us to successfully optimize the approximation process [1,2]. A feature of the optimization algorithm is the need to configure the internal parameters of the algorithm and the strategy for selecting mutant vectors that affect the speed of convergence. Using simulated spectra as experimental spectra, we analyzed the stability of the algorithm. It turned out that for some differential evolution strategies the parameters ensuring fast convergence are in a fairly narrow range. Thus, preliminary testing of strategies is necessary before the algorithm processes real experimental data.
References 1. Pishchalnikov, R.Y.; Yaroshevich, I.A.; Zlenko, D.V.; Tsoraev, G.V.; Osipov, E.M.; Lazarenko, V.A.; Parshina, E.Y.; Chesalin, D.D.; Sluchanko, N.N.; Maksimov, E.G. The role of the local environment on the structural heterogeneity of carotenoid β-ionone rings // Photosynthesis Research, том 156, номер 1, 2022, Стр. 3-17, doi:10.1007/s11120-022-00955-2. 2. Pishchalnikov, R. Application of the differential evolution for simulation of the linear optical response of photosynthetic pigments // Journal of Computational Physics том 372, 2018, Стр. 603-615, doi:10.1016/j.jcp.2018.06.040.
Presentation |
