Structural Optimization of Spoke Single-Layer Cable-Net Structures Based on a Genetic Algorithm

Abstract

A structural optimization method for spoke single-layer cable-net structures based on an improved genetic algorithm (GA) is proposed. The three-cable coplanarity theory is introduced, and an exchange program between a proprietary programming language and an advanced finite-element software is presented to implement the improved GA. Then, a simplified numerical model of the Suzhou Industrial Park Stadium cable-net structure is built and optimized. The numerical example shows that the proposed structural optimization method is accurate and efficient. Because of the limited loading combinations, nodal vertical displacements, and tensile stresses of cables, the optimal z coordinates of support nodes and cable nodes and subsequently, the initial pretension and cross-sectional areas of cables, are obtained. When the z-coordinate distribution curve of the support nodes behaves smoothly, the material property of the cables will be fully utilized with a reduction in cable weight. For the optimal mode of the simplified model, the elevations of the support nodes provide the entire cable-net structure a saddle shape, which behaves rigidly under out-of-plane loads. The proposed method can be extended to solve optimization problems of similar cable-net structures.