基于神经网络的结构边界条件识别和损伤诊断

结构边界条件识别和损伤诊断，对于实验室条件下研究结构的动力特性具有重要的实用价值，本文提出了以子网为基础的分区组合式神经网络模型，并以钢筋混凝土梁在静力作用下的挠度作为特征参数，实现了结构的边界条件识别和损伤诊断。

Structural Boundary Conditions Recognition and Damage Diagnosis Based on Neural Networks

Structural boundary conditions recognition and damage diagnosis are of important practical value for the studies of strutural dynamic properties in laboratory. This paper presents a kind of partition combination neural network model based on subnets. Structural boundary conditions recognition and damage diagnosis are realized by the use of static deflections of reinforced concrete beam as characteristic factors. The results indicate this method is feasible.