|
|
|
|
|
|
|
Description
Anthropometry is the science of human measurement. The
role of anthropometry in engineering is to provide measurements and methods for
using human measurement to design products and services. The tutorial is based
on recently developed techniques using 3D body scanners and modeling software
for the achievement of effective anthropometry in engineering.
Outline
Abstract
Evolutionary computation (EC) has been recognized as a current research
field, which studies a new type of algorithms: Evolutionary Algorithms (EAs).
These algorithms process populations of solutions as opposed to most traditional approaches, which improve a single solution. All these algorithms share common features: reproduction, random variation, competition and selection of
individuals.
During our research it was evident that some components of EAs should be
re-examined. Exploration and exploitation of solutions in the searching space
are distinctive characteristics of an evolutionary algorithm, and are
responsible for the success or failure of the search process. Extreme
exploitation can lead to premature convergence and intense exploration can make
the search ineffective. To find a balance between these two factors is of
paramount importance for the EA performance when speed of the search and quality
of results are involved. Many researchers focus on the balancing problem
studying the effect of selection mechanisms, because selective pressure can
adjust exploration and exploitation. On its own, recombination can also
participate on this respect but depending on how it is applied it can aid or
disrupt the search process. For example, a low rate for recombination can impede
schema processing permitting super-individuals to replenish the population, thus
leading to premature convergence. On the other hand a high rate can be, in some
cases, extremely disruptive allowing good genetic material to be lost, slowing
down the search.
Parallel implementations of Evolutionary Algorithms also aim at improvements
on performance. The main purpose of this approach is to enhance the quality
of the results. The island model, a well known distributed approach, where
separate subpopulations evolve in parallel is a realistic model of natural evolution
which is appropriate for a distributed environment running a Single Program
Multiple Data (SPMD) scheme.
This tutorial will show the most relevant and recent enhancements on recombination for genetic-algorithm-based EAs and migration control strategies for parallel genetic algorithm applied to diverse optimization problems including scheduling.