【摘要】 针对单点损伤识别方法的缺点,提出了基于频响函数和小波包能量谱的斜拉桥多点损伤识别指标。对结构损伤前后的频响函数值进行了随机-模糊均值处理,将处理的频响函数作为基进行小波包能量谱分析,提取能量累积变异值进行损伤识别,并进行了独塔斜拉桥模型的试验研究和连续梁与简支梁的数值仿真。分析结果表明:桥梁上有1处和2处损伤时都可被识别和定位;损伤指标随着损伤程度的增加而变化,损伤初期呈线性变化,损伤程度在40%~50%时,指标变化趋缓,损伤超过50%后,指标值又会急剧变化;采用锤击激励比环境激励的损伤识别结果更好,锤击激励力大小对损伤识别效果无影响。更多还原

【Abstract】 In view of the defects of single-location damage identifying method,the multi-location damage identifying index of cable-stayed bridge in combination with wavelet packet energy spectrum and frequency response function(FRF) was proposed.The random-fuzzy mean was processed with the FRFs of undamaged and damaged structures.Wavelet packet energy spectrum was analyzed with the processed FRFs as wavelet base functions,and the value of accumulated energy variation was extracted to recognize damages.The experiment of a single tower cable-stayed bridge model and the numerical simulations of continuous and simple-supported beams were carried out.Analysis result shows that 1 or 2-loacation damage of bridge model can be identified and located.With the increasing of damage severity,the initial damage index increases linearly,and it will change gently or dramatically when the damage severity approaches 40%-50% or exceeds 50% respectively.The damage identifying result of hammering excitation is better than that of environmental excitation,and the value of hammering exciting load has no effect on identifying precision.1 tab,8 figs,15 refs.更多还原