08.06.2017 Views

Application of Genetic Algorithm in Multi-objective Optimization

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

1. Introduction:<br />

<strong>Genetic</strong> algorithms (GAs), members <strong>of</strong> the large class <strong>of</strong> “Evolutionary <strong>Algorithm</strong>s”, are<br />

metaheuristics approach for solv<strong>in</strong>g various optimization problems. GAs are <strong>in</strong>spired by the natural<br />

selection process. GAs operate on a set <strong>of</strong> potential solutions apply<strong>in</strong>g the pr<strong>in</strong>ciple <strong>of</strong> survival <strong>of</strong><br />

the fittest to produce better and better approximations to a solution. Based on the fitness level <strong>of</strong><br />

the <strong>in</strong>dividual solution <strong>in</strong> each generation, a new pool <strong>of</strong> parents is selected for breed<strong>in</strong>g the next<br />

generation us<strong>in</strong>g various operators adopted from natural evolution. Thus at each generation, GAs<br />

try to generate <strong>of</strong>fspr<strong>in</strong>g exhibit<strong>in</strong>g better fitness level which are better suited to their environment<br />

than the population they are generated from [1]. In various multi-<strong>objective</strong> optimization problems,<br />

the applicability <strong>of</strong> GAs has been proven through numerous research works. Also, there are<br />

compet<strong>in</strong>g optimization methods available to solve structural design problems. But certa<strong>in</strong><br />

characteristics <strong>of</strong> this class <strong>of</strong> problems have made GAs popular <strong>in</strong> this research field. GAs are<br />

suitable for cont<strong>in</strong>uous problems as well as for discrete and non-differentiable problems.<br />

Additionally, GAs are very efficient for search<strong>in</strong>g global optimal solutions.<br />

However, there are some <strong>in</strong>terest<strong>in</strong>g areas related to the application <strong>of</strong> GAs to the structural<br />

optimization problems which are not yet fully explored. The follow<strong>in</strong>g areas require further<br />

<strong>in</strong>vestigation<br />

<br />

In constra<strong>in</strong>ed optimization problems, application <strong>of</strong> penalty function is very common. GAs are<br />

successfully used with penalty function applications. However, there is no systematic approach<br />

to understand the <strong>in</strong>fluence <strong>of</strong> the magnitude and trend <strong>of</strong> penalty function on the convergence<br />

towards the global optima.<br />

14

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!