基于无线传感与稀疏时频分析的桥梁拉索时变索力识别

发布时间：2018-04-24 20:00

本文选题：结构健康监测 + 稀疏自适应的时频分析方法　；参考：《哈尔滨工业大学》2015年硕士论文

【摘要】：疲劳累积损伤是实际桥梁结构发生损伤破坏的主要原因之一,对于缆索类桥梁,拉索是其关键受力构件,拉索的累积疲劳损伤严重威胁桥梁结构的安全。传统的基于振动法的索力测试,只能识别桥梁拉索在一段时间内的索力平均值。然而,桥梁拉索由于车辆荷载和环境因素的作用,其索力是时变的。时变索力是引起疲劳损伤的主要原因,同时也是进行拉索极限状态安全评定和疲劳累积损伤评价的基础。为此,本文研究基于无线传感器和稀疏时频分析的桥梁拉索时变索力识别方法。本文主要研究内容包括:基于稀疏自适应的时频分析方法,提出一种新的识别时变索力的方法。稀疏自适应的时频分析方法是当前信号处理领域新发展的方法,其原理是通过在最大的时频分析字典里优化寻找信号的最稀疏分解得到该信号的瞬时频率。首先,根据稀疏自适应的时频分析方法,从测得的索的加速度识别出时变的模态频率。然后,将桥梁拉索简化为理想的张紧弦,依据经典弦振动理论建立的索力与频率之间的关系进行索力识别。利用索的振动存在的倍频关系,减少方法的优化变量,提高计算效率、降低噪声对索的不同阶瞬时频率的影响、提高时变模态频率识别的精度。最后将识别得到的索的时变模态频率,代入索力公式得到拉索的时变的索力。采用拉索模型的有线传感器试验数据,考虑不同的索力变化工况,研究索力变化水平和速度对于本文提出方法识别精度的影响。然后考虑采用无线传感器可能产生的数据丢失问题,利用有线传感器测得的加速度数据模拟可能发生的数据丢失的不同情况,基于丢失后的数据进行索力识别,研究数据丢失的模式和数据丢失率对于该改进的时频分析方法索力识别结果的影响。研究采用Imote2无线传感器进行拉索振动测试和索力识别,对拉索模型进行试验并考虑不同的索力变化工况。由于无线传感器易受外界环境干扰而产生数据丢失,在试验条件下人为造成数据丢失,对没有丢失和存在数据丢失的实验数据分别进行稀疏自适应的时频分析,根据时频关系识别索力,并对比两种情况下识别到的频率以及索力,测试稀疏自适应的时频分析方法在存在数据丢失的情况下对索力的识别的精度和鲁棒性。并对厦门海沧大桥的吊索进行现场测试,利用Imote2无线传感器,测试实桥吊索的振动加速度,采用论文提出的方法识别时变的模态频率并计算时变索力,进一步验证方法对实际桥梁拉索时变索力的识别效果。
[Abstract]:Fatigue cumulative damage is one of the main causes of damage and damage to the actual bridge structure. For cable bridges, the cable is the key component of the bridge. The cumulative fatigue damage of the cable is a serious threat to the safety of the bridge structure. The traditional cable force test based on the vibration method can only identify the cable force average of the cable in a period of time. The cable force of bridge is time-varying due to vehicle load and environmental factors. The time variant cable force is the main cause of fatigue damage, and it is also the basis of safety assessment and fatigue cumulative damage evaluation of cable. Therefore, this paper studies the time variation of bridge cables based on wireless sensor and sparse time-frequency analysis. The main contents of this paper are as follows: Based on the sparse adaptive time-frequency analysis method, a new method of identifying time variant cable force is proposed. The sparse adaptive time-frequency analysis method is a new development method in the field of signal processing, and its principle is to optimize the thinnest of finding the signal in the largest time-frequency analysis dictionary. The instantaneous frequency of the signal is obtained by sparse decomposition. First, according to the sparse adaptive time-frequency analysis method, the time-varying modal frequency is identified from the acceleration measured by the cable. Then, the bridge cable is simplified to an ideal tension string, and the cable force identification is carried out according to the relationship between the cable force and the frequency rate based on the classical string vibration theory. In order to improve the calculation efficiency, reduce the influence of the noise on the different instantaneous frequency of the cable and improve the accuracy of the time-varying modal frequency identification, the time-varying modal frequency of the identified cable is obtained, and the cable force of cable is obtained by using the cable force formula. The cable sensor using the cable model is used. The test data, considering the different cable force change conditions, study the influence of the cable force change level and speed on the method recognition accuracy. Then consider the data loss problem that the wireless sensor may produce, using the acceleration data obtained by the wired sensor to simulate the different data loss, based on the loss. The later data are identified by cable force, and the effects of data loss patterns and data loss rates on the results of the improved time-frequency analysis method are studied. Imote2 wireless sensor is used to test the cable vibration and cable force identification, and the cable model is tested and different cable force changes are taken into consideration. Data loss is caused easily by external environment interference. In experimental conditions, the data is lost, and the time-frequency analysis of the experimental data which is not lost and the data loss is sparse adaptive. According to the time frequency relation, the cable force is identified and the frequency and cable force identified in the two cases are compared, and the time frequency of the sparse adaptive time is tested. The analysis method is accurate and robust to the identification of cable force in the case of loss of data. In the field test of the sling of Xiamen Haicang Bridge, the vibration acceleration of the real bridge sling is tested by Imote2 wireless sensor, and the time-varying mode frequency and time variable cable force are identified by the method proposed in this paper, and the method is further verified. The identification effect of the variable cable force when the actual bridge is pulled.

【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U446

【参考文献】