Structural analysis of GFRP elastic gridshell structures by particle swarm optimization and least square support vector machine algorithms

The gridshell structure is a kind of freeform structure, which is formed by the deformation of a flat grid and the final shape is a double curvature structure. The structural performance of the gridshell is usually obtained by finite element analysis (FEA), which is a time-consuming procedure. This paper aims to present a framework for structural analysis based on the machine learning (ML) model in order to reduce computational time. To this aim, design parameters including the length, width, height, and grid size of the structure are taken into consideration as inputs. The outputs are the member-stresses and the ratio of displacement to self-weight. Therefore, a combination of two algorithms, least square support vector machine (LSSVM) and particle swarm optimization (PSO), is considered. PSO-LSSVM hybrid model is applied to predict the results of the structural analysis rather than the FEA. The results show that the proposed hybrid approach is an efficient method for obtaining structural performance.