动载荷下梁结构非线性损伤的光纤光栅实时监测和识别方法

The early nonlinear damage diagnosis of engineering structures is now a challenging problem in the field of structural damage identification. The research program of this project is suggested based on the practical problem that the existing electrical sensing technology very difficultly realize to real-timely monitor and extract the characteristic signal of the early nonlinear damage of large and complex structures on the long distance and distributed scale. An experimental system has been constructed according to the finite element analysis results in order to real-timely monitor the health status of metal beam structures using FBG strain sensor array and a fiber grating vibration sensor array. The research works of this project are included: (1) to prefabricate the early small cracks using the laser etching method; (2) to investigate the influences of the early crack characteristics (shape, size, and quantity) on the stress strain and intrinsic mode parameters of metal beam structures under condition of a constant load and a fixed vibration frequency; (3) to investigate the fiber grating real-time monitoring method of the stress strain and intrinsic mode parameters of metal beam structures under dynamically loading and random vibration frequency; (4) to reveal the nonlinear variation laws of the nonlinear damage of beam structure under dynamically loading induced by the early crack growth; (5) to extract the nonlinear damage characteristic signal induced by the early crack growth by employing the wavelet transform time-frequency analysis method; (6) to establish a model for structural damage BP neural network intelligent identification. The targets of this project are as followings: (1) to develop a new identification method for the nonlinear damage of beam structures under dynamically loading based on the fiber grating real-time monitoring; (2) to promote the application of the long-distance, distributed, multi-parameter fiber grating sensing technology in the early nonlinear damage identification of large and complex structures.

工程结构早期非线性损伤诊断是目前结构损伤识别领域的挑战性难题。针对现有电类传感技术难以实现远距离、分布式、实时监测和提取动载荷下大型复杂结构早期非线性损伤特征信号问题，本项目提出搭建金属梁结构健康状态的光纤光栅实时监测实验系统，采用激光预制早期微小裂纹，通过光纤光栅应变传感阵列和光纤光栅振动传感阵列并用，研究恒定载荷下早期裂纹特征(裂纹大小、数量、形状)对金属梁结构应力应变和本征模态参数的影响规律；在此基础上，重点研究动载荷下早期裂纹扩展过程中金属梁结构应力应变和本征模态参数的变化以及裂纹扩展引起的金属梁结构非线性损伤演变规律；采用小波变换时频分析方法提取早期裂纹扩展产生的非线性损伤特征信号，构建结构损伤BP神经网络智能识别模型。建立动载荷下梁结构非线性损伤的光纤光栅实时监测和识别新方法，推动光纤光栅远距离、分布式、多参数传感技术在大型复杂结构早期非线性损伤的实时监测与识别中的应用。

针对工程结构早期非线性损伤诊断面临的困难以及电类传感器在实时监测中存在的不可克服的缺陷，本项目以建立动载荷下梁结构非线性损伤的光纤光栅实时监测和识别方法为目的，开展了金属悬臂梁结构和金属简支梁结构的有限元分析、基于光纤光栅应变传感阵列和光纤光栅振动传感阵列的损伤简支梁的动态监测与识别以及动载荷下裂缝简支梁非线性损伤的光纤光栅传感阵列识别方法的研究工作。取得了以下主要研究成果：.(1).建立了金属悬臂梁几何非线性的有限元分析模型，采用有限元分析软件模拟了金属悬臂梁结构在受到固定载荷情况下的应力应变分布状态，获取了金属悬臂梁结构损伤前后的应变变化规律，确定了金属悬臂梁承受固定载荷产生几何非线性的最小载荷，获得了金属悬臂梁结构在损伤状况下的应力-应变关系和损伤大小对金属悬臂梁固有频率变化的影响。和动载荷作用下.(2).建立了基于分布式光纤光栅传感原理的悬臂梁和简支梁应变检测方法，研制出光纤光栅应变传感阵列并研究了传感阵列的静态特性，搭建了光纤光栅传感阵列实时监测金属悬臂梁结构和简支梁结构损伤的实验系统，采集并分析了光纤光栅检测的金属悬臂梁损伤前、后的应力应变数据以及简支梁在振动状况下的实验数据，揭示了振动状况下金属简支梁结构损伤后产生的非线性信号变化规律和0-200Hz恒定动载荷条件下金属梁结构的应变和频率变化规律。.(3).采用小波变换时频分析方法提取了金属简支梁结构的未损伤特征信号和损伤特征信号，推导了基于BP神经网络的小波神经网络算法，建立了基于光纤光栅传感应变阵列和光纤光栅振动传感阵列测量的简支梁损伤识别的BP神经网络模型和基于BP神经网络与光纤光栅传感阵列的简支梁损伤识别方法，研制出动载荷下检测金属梁结构非线性损伤的光纤光栅传感阵列实时监测实验系统。.这些研究成果将推动光纤光栅远距离、分布式、多参数传感技术在大型复杂结构早期非线性损伤实时监测与识别中的应用。