Modified bidirectional evolutionary structural optimization for continuum structures

Recently, some researchers proposed a new algorithm for the bidirectional evolutionary structural optimization (BESO) in which adding and removing of elements is controlled by a signal parameter of removal ratio of volume (or weight). Nevertheless, their BESO improvement suffers from some restrictions like premature stopping of the iterative process due to the closeness of the adding and removing coefficients. In this study, the BESO method is modified such that eliminates the previous restrictions and increase its efficiency significantly. Several examples are presented for structural optimization of a plate in two states of elastic and elastoplastic with stiffness and stress criteria. From the three examples, one can conclude that the modified BESO algorithm is more robust than the ESO and BESO algorithm previously proposed. It is also concluded the maximum stress and the performance index are improved in comparison with the case when the signal parameter is the same. Overall, the proposed modification prevent the premature termination of the optimization process when the optimization process is based on the hard kill method with both adding and removing the elements.