engleski

Evolutionary Algorithm Adapted for Matching Weave and Color in Fabrics Design

Application of evolutionary algorithms for fabric design imposes numerous limitations, especially in the population size and the number of fitness evaluations. Fitness evaluation for matching weave and colour in fabric design is a human aesthetic evaluation, so it is necesarry to take into account psychological aspects. Standard evolutionary algorithm has on average the population size 50 - 100, the number of generation 500 - 5000, which leads to 25000-500000 fitness evaluations. Because of temporal and psychological limitations, such algorithms are not appropriate for the applications in which fitness evaluation is a human subjective evaluation. For these reasons, instead of standard evolutionary algorithms, interactive micro evolutionary algorithms are used. Their characteristics are relatively small population size (8-15) and a small number of generations (5-15). Interactive micro evolutionary algorithm is described and its application for matching weaves and colors in fabrics design proposed. Its acceptability for human subjective evaluation is also investigated.

Interactive evolutionary algorithm; weave and colour; micro evolutionary algorithm; fabric; MATLAB

nije evidentirano