OPTIMUM DESIGN OF TRUSSES USING GENETIC ALGORITHMS

Abstract

The goal of structural engineers is to design structural systems according to certain design requirements. But, having a system that just performs the required task satisfactorily in not sufficient. It is essential that the design must be the best or optimal based on the specified requirements. An optimal design should be a cost effective system. To design such systems, certain analytical and numerical tools are needed. Optimal design concepts and methods provide some of the necessary tools. The Genetic Algorithms which are the search and optimization algorithms that mimic the natural evolution and genetics are potential optimization algorithms, and have been applied to many engineering design problems in the recent past. Due to their population approach and parallel processing, they have been able to obtain optimal solutions in complex optimization problems. The weight optimization of trusses, winch is a discrete optimization problem, can not be solved effectively using conventional optimization techniques. The genetic algorithms are well suited for discrete optimization. In the present study, the truss structures are optimized using the Genetic Algorithm (GA), which belong to the area of 'Evolutionary Computation' and 'Soft Computing'. The effect of various GA parameters is studied in depth to find out the range of GA parameters which makes GA to perform well. The effect of selection pressure is also studied. The effect of number of design variables on optimum achieved is also investigated. The stress constraint for compression members, with and without buckling effect is applied. In total three problems, two 2-dimensional trusses and one space truss are optimized. In all the cases, GA gives results which are better than the results available in literature.