Adaptive Step-wise Bi-directional Evolutionary Structural Optimization Method

In the process of structural optimization design for components, fixed step size in the Evolutionary Structural Optimization and Bi-directional Evolutionary Structural Optimization methods result in slow convergence speed and excessive element deletion. To resolve these issues, we propose an adaptive-step BESO method. It employs discrete performance indicators and control factors to dynamically adjust step sizes, aiming to minimize the harmonic response of nodal displacements. By accounting for variations in structural volume and element sensitivity, this method improves convergence rate and stability. The resulting topology meets volume constraints without empirical estimations or oscillations. Numerical examples confirm that the adaptive-step approach enhances computational efficiency, achieves progressive convergence in volume and harmonic response, and produces a clear, well-defined topology.