Damage Identification Method of Space Truss Based on Elemental Strain Mode Difference and Neural Network

摘要

Now damage identification on space truss became difficult to realize because of these structures with multiple links, high stiffness and complex structural shapes. To solve these questions, a method combined with elemental strain mode difference and neural network was proposed to research damage identification of the space truss structures in this paper. Firstly, the damage locations were determined by elemental strain mode difference which came from the elemental strain mode differences of damaged and intact space truss. Secondly, the damage degrees of this structure were discerned using RBF neural network which established by the elemental strain mode difference vectors of damaged and intact structure. At last, the damage locations and damage degrees of a square pyramid space truss were implemented by this method. The results show that this method can be used to detect the damage locations and define its damage degrees.