Program as well as the associated estimation method, the substance structure based optimization method of multiple parameters by the hybrid GA, data analysis method by pattern matching, the radiation diffraction data, various devices and recording media

摘要

A hybrid genetic algorithm (HGA) in which local optimization is conducted on only an extremely small number of individuals showing a relatively low fitness in a number of generations widely distributed in genetic algorithm (GA). Because of exhibiting an extremely strong ability to search for over a wide range even in a multidimensional parameter space having an intense multi-peak property, this HGA makes it possible to efficiently determine a large number of structural parameters from little information involved in an X-ray diffraction ring. As a result, a complicated structure of a substance can be determined by using a powdery sample without resorting to a single crystal, which accelerates the development of novel substances in many fields including medicines and raw materials.