OPTIMUM PLASTIC DESIGN OF FRAMED STRUCTURES IN CONSIDERATION OF ROBUSTNESS BY NONLINEAR SENSITIVITY : Application of parallel GA using PC cluster to multi-objective optimization

Abstract

This paper presents robust plastic design for plane frames which are subjected to horizontal and vertical loads. The purpose of the robust design is to minimize of weight and nonlinear sensitivity of the collapse loads by the strength of members. It is desirable to design the structures that the collapse load is only slightly modified by variations of the strength of members. The multi-objective optimization problems are solved using parallel genetic algorithms(PGA). Computational experiments are performed on PC(Personal Computer) cluster using PVM(Parallel Virtual Machine) which is a software package. The pareto-optimal solutions found by PGA were confirmed by using Monte Carlo simulation.