Truss sizing and shape optimization with stress and local buckling constraints

摘要

A typical constrained mixed-integer nonlinear program: truss sizing and shape optimization under multi-loading cases and complex constraints is investigated in this work. A laconic adaptive penalty function is employed to handle the structural stress and local buckling constraints. The Covariance Matrix Adaptation Evolution Strategy (CMA-ES) is adopted to update the structural design variables. The effectiveness and efficiency of proposed hybrid algorithm is verified by a benchmark tower optimization problem. Numerical results indicate that the performance indicators of present work are better than those shown in the literature.