Artificial-Neural-Network-Based Mechanical Simulation Prediction Method for Wheel-Spoke Cable Truss Construction

Cable force monitoring is an important step in cable truss structural health monitoring. Considering cost eff ectiveness, the accuracy and quality of safety assessments depend primarily on the usage of reasonable cable monitoring programs. Many monitoring methods have been proposed to design the cable truss structure. The emergence of artifi cial neural network (ANN) models has resulted in improved predictive abilities. In this study, an ANN-based model is used to estimate parameter changes in static and prestress loss tests during the construction of a cable truss. The fi nite element model data of 243 cases are analysed by ANSYS and MATLAB. Analysis results indicate the excellent prediction performance as well as high accuracy and generalization of the proposed ANN-based model. Furthermore, the successfully trained ANN-based model is used to predict new cases. As an alternative to fi nite element analysis and physical test, the proposed model can guide the static loading of the spoke cable truss structure and thus allow the safe usage of the structure during service. KCI Citation Count: 0