Optimal design of truss structures via an augmented genetic algorithm

This paper investigates the application of an augmented genetic algorithm (AGA) for the problem of optimal design of truss structures. The problem is to determine the minimum value of the weight/cost associated with the truss structure design while a set of stress and displacement constraints are to be satisfied. The proposed AGA exploits a probabilistic selection procedure based on the annealing process. Moreover, a new enhancing trick is proposed that prevents familial reproduction. This effect restrains the degenerative phenomena during the evolution of the canonical genetic algorithm (GA). Accordingly, a considerable improvement in converging speed is achieved. This advancement could be extremely advantageous in large structure design problems, where the existing methods suffer from long execution times. Various benchmark examples are examined to demonstrate the performance of the proposed method. Moreover, the obtained results are compared with those of existing methods to verify the effectiveness of AGA optimization.