USING EVOLUTIONARY ALGORITHM AS A DESIGN TOOL FOR THE MULTI CRITERIA OPTIMIZATION OF CATENARY STRUCTURES

Abstract

The aim of this paper is to demonstrate the use of Evolutionary Algorithm to optimize catenary models. The processes of optimizing structures using the conventional model of physically hanging strings and weights are long and tedious. The resultant model forms a catenary shape and hence is called a catenary model. During physical modeling experiments, vertical forces of the self-weight of the model and the predicted imposed loadings are taken into account.

Evolutionary Algorithms is a method which is able to automate the above process. It is a set of rules, which follows the process of natural selection, in order to solve problem within a finite number of steps. This set of rules contains the parameters which dictate different variances of the catenary models. The natural selection comprises of evaluation criteria necessary for the selection of a successful catenary model. The advantage of applying this method is the ability to include horizontal forces, such as wind load.

This paper will (i) introduce what Evolutionary Algorithm is and how it is applied; (ii) give a detailed outline of the experiment which demonstrates the use of Evolutionary Algorithm to optimize catenary models; (iii) problems faced using this method, future explorations of this topic and conclusion.