Algorytmy ewolucyjne w projektowaniu fraktalnych wzorów użytkowych

Abstract

The aim of the PhD thesis is an automatic search of fractal patterns with the highest value of aesthetics. Nowadays designers pay a lot of attention to product design. Consumer market demand is associated with not only functionality of the product, but also with its aesthetics. Expectations of the consumer while using the product, which may be an electronic device or software, are described as “user experience” (UX). It involves the interaction between man and computer, and refers to the functionality, comfort and work satisfaction. The design of the product has among others the impact on these factors. In the work genetic algorithms were used, which require just direct search in the set of feasible solutions, using estimation of the evaluation function of individuals - fractal patterns. The evaluation function was based on the statistical research on group of 100 respondents. Experimental studies using different types of genetic operators, selection and reproduction of individuals were presented in the PhD thesis. The effect of genetic parameters of the algorithm for its convergence was examined. Finally geometric patterns with a nontrivial structure and aesthetic values were generated, which accounted for the purpose of research. The PhD thesis includes defining the characteristics and algorithm which allows fully automatic generation of aesthetic fractal patterns, according to the valuation of these characteristics, imitating, in simple terms, human perception. The research includes the study of fractals and dynamic systems in terms of visual, formalization of important from the point of view of human perception characteristics that describe the fractal structure, and determining their parameters. Using these parameters, it is possible to identify from a large number of fractal patterns structures showing its construction, for the average person, beauty.