基于曲率模态和神经网络的施工桁架梁损伤识别研究

Damage Identification of Construction Truss Beam Based onCurvature Mode and Neural Network

桁架梁在大跨施工中作为组装式承重结构,代替高支模体系作施工平台,关系到人员生命财产安全,需要进行损伤检测。在人工观察法检测基础上,为提高识别准确率和效率,提出基于曲率模态和神经网络的损伤识别两步法。结合西安某产业研发基地在役施工桁架梁,通过ABAQUS分析模态参数,使用小波变换对数据进行处理,实现损伤定位,并构造BP神经网络,以识别损伤程度。研究表明,基于曲率模态及小波系数和BP神经网络的损伤识别两步法可较好实现对桁架梁的损伤识别。

Truss beam is used as assembled load⁃bearing structure in long⁃span construction,replacing high formwork system as construction platform.It is related to the safety of human life and property,so it is necessary to detect its damage.Based on the existing manual observation method,a two⁃step damage identification method based on curvature mode and neural network is proposed to improve the identification accuracy and efficiency.Taking the in⁃service construction truss beam of an industrial research and development base in Xi’an as an example,the model parameters are analyzed by ABAQUS,and the wavelet transform is used to process data to realize the damage location,and the BP neural network is constructed to realize the damage degree recognition.The research shows that the two⁃step damage identification method based on curvature mode,wavelet coefficient and BP neural network can better realize the damage identification of truss beam.