桥梁结构动态应变监测信息的分离与提取

对工作应变传感器和温度应变传感器的数据进行了对比分析,利用能量谱比指标将监测系统输出的原始应变数据按照温度、应力和测量噪声3种类型在频域中进行分类,分别确定了温度变形与结构应变的分界频率.提出了一个多分辨递阶方法,可将3类不同信息分离并提取出由结构应力导致的应变信息,在分离信息基础上考察了干扰部分和温度应力部分对疲劳评估的影响.研究结果表明,2个分界频率值对同类桥梁有较强的参考价值,提出的算法具有多分辨和递阶的特征,适合于海量应变数据的压缩与预处理,便于在结构健康监测系统软件中集成.虽然变温变形部分的能量很大,但其对疲劳评估产生的误差很小;而干扰信息虽然能量很小,但对疲劳评估产生的影响很大,其主要原因是干扰信息使得应力应变中提取的应变幅发生偏移.

Separation and extraction of bridge dynamic strain data

Through comparing the measured data of dynamic strains due to loading and temperature by the strain gauge and temperature sensor at the same location,the information in the strain data was divided into three parts in the frequency domain by using the defined index named PSD(power spectral density)-ratio.The three parts are dominated respectively by temperature varying,stresses and noises and can be distinguished from the determined values of the separatrix frequencies.Then a simple algorithm was developed to separate the three types of information,and to extract the strain caused mainly by structural stresses.As an application of the proposed method,the influence of strain deformation and noises on the fatigue assessment was investigated based on the separated data.The results show that,the determined values of separatrix frequencies are valuable for the monitoring data from other bridges.The algorithm is a multi resolution and hierarchical method,which has been validated as a simple and effective method for data analyses,and is suitable for the compression and preprocessing of the great amount monitoring data and easy to be integrated into the SHM's(structural health monitoring) software system.The strain due to temperature varying attributes only a little to the errors of fatigue assessment,however,the noises or random disturbance existed in the monitoring data have much responsibility for the errors,the main reason is that the random disturbance shifts the real strain/stress amplitude picked up by real structural stress or strain.