A Genetic Evolution Algorithm for Structural Optimization

摘要

The application of genetic algorithm-based methodology for the structural design is presented in this study. The genetic algorithm is used to design prestressed concrete beams (PCB). The target objective in this method is to obtain set of optimal geometrical dimensions of symmetrical I-beam cross section. Additionally, the amount of pre-stressing steel is optimized. Post-tensioned prestressed beam with a single duct of parabolic shape is considered in the application. Several parameters are studied including the effect of the span length considering different loading cases. The performance constraints are adopted according to the ACI 318/95 Building Code provisions [1]; including the flexural stresses, the ultimate moment capacity of the section with respect to cracking moment, the maximum crack width, the immediate deflection and the long term deflection in addition to the side constraints. The results are presented and compared; several design charts are developed and presented. The present study showed the promising capabilities of the genetic algorithm in optimal designs, and showed the practicability of the genetic algorithm for different structural optimization problems.