PHYSICS-INFORMED DIFFERENTIAL EVOLUTION FOR NONLINEAR ANALYSIS OF CABLE NETS

Abstract

In this article, a physics-informed differential evolution approach is developed to estimate the nonlinear behaviors of pretensioned cable structures without using any structural analyses. Rather than solving nonlinear equations through conventional numerical methods, this approach employs differential evolution to minimize total potential energy (TPE). Therein, the TPE is designed as an objective function to guide the searching process of the differential evolution (DE) algorithm. Once the minimum TPE value is found, the nonlinear behavior of structures can be easily obtained. Three benchmarks are examined to determine the efficiency of the suggested framework for geometrically nonlinear analysis of cable structures. The results demonstrate that the proposed approach is easy to implement and delivers high accuracy.