Premium
Application of immune and genetic algorithms to the identification of a polymer based on its X‐ray diffraction curve
Author(s) -
Rabiej Małgorzata
Publication year - 2013
Publication title -
journal of applied crystallography
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.429
H-Index - 162
ISSN - 1600-5767
DOI - 10.1107/s0021889813015987
Subject(s) - diffraction , algorithm , crystallinity , genetic algorithm , amorphous solid , polymer , materials science , set (abstract data type) , identification (biology) , x ray crystallography , computer science , optics , crystallography , physics , mathematics , chemistry , mathematical optimization , composite material , programming language , botany , biology
This paper describes how a combination of two methods of artificial intelligence, an immune algorithm and a genetic algorithm, can be used to recognize a polymer by the shape of its X‐ray diffraction curve. To this end, the hybrid algorithm uses a database which contains theoretical functions describing wide‐angle X‐ray diffraction curves of different polymers. These curves are compared by the algorithm with the experimental diffraction curve and the most similar are chosen. Such theoretical curves are kept in the immunological memory, and their parameters can be set as the starting ones in the optimization methods used for decomposition of the experimental curve into crystalline peaks and amorphous component. Using this algorithm, the preparation of the starting parameters is much easier and faster. Decomposition is the most important step in polymer crystallinity determination.