Surrogate-assisted multi-objective topology optimization design method for prefabricated large-span Suspen-Dome based on structural-mapping-method and improved-PGSA

Existing optimization studies on large-span Suspen-Dome structures have mainly focused on either size optimization of the upper reticulated shell or topology optimization of the lower cable-strut system. Such a separated strategy prevents the integrated optimization of the whole structure, restricts the design space, and limits further improvement in structural performance. This study investigates the integrated optimization of prefabricated large-span Suspen-Dome structures by simultaneously considering the size, geometric configuration, and topology of the upper reticulated shell, together with the topology and cable forces of the lower cable-strut system. To ensure practical applicability, the design variables are classified into topological and size variables. The former define the structural configuration and connectivity, while the latter determine material-related properties. The optimization aims to maximize stiffness and minimize weight, subject to member slenderness, material strength, local stability, and overall structural stability constraints. An Multi-Layer Perceptron (MLP)-based candidate screening model is introduced to accelerate the optimization algorithm and reduce unnecessary ANSYS analyses. A partial full-stress design (PFSD) concept is proposed, in which the reticulated shell serves as a structural safety reserve. The proposed structural mapping method-improved plant growth algorithm (SMM-iPGSA) is validated through two examples. Results show that the optimized structures exhibit improved stability, more uniform stress distribution, and displacements well below the allowable limits.