Multiobjective optimization of bar structures by Pareto-GA

Abstract

© ECCOMAS.The optimisation problem considered here is to minimize the weight of frames with discrete real cross section types under a multiobjective context: • First, the minimisation of the weight of the structure taking into account different constraints. • Second, the minimisation of the number of different cross section types, important in large structures for constructive requirements. The constraints are: stresses of the bars, displacements of joints or middle points of bars, and slenderness limits to include the buckling effect (as described in the Spanish code). We solve the minimisation problem using an elitist strategy of the Non-dominated Sorting Genetic Algorithm (NSGA) and Gray Code. Results are compared with different simple objective GA strategies, such as generational, steady-state or CHC, and the results obtained in different applications demonstrate that the maintenance of partial solutions in the front by the NSGA is carried out with successfully evolution towards to the Pareto solution. The introduction of the elitist operator improves significantly the quality of the obtained solution and inclusion of high quality solutions in the initial population can be also a way to obtain improved final fronts.