A Study On Data Processing Method And Software Development Of Bridge Health Monitoring

As the problem of the safety of bridge getting more and more crucial, the traditional manual detecting method cannot guarantee the operation safety of the long-span bridge any more. Thus, the bridge health monitoring system has gradually become the developing direction for the bridge operation and maintenance. However, the bridge health monitoring system nowadays emphasizes the hard-wares and data-gathering while neglects soft-wares and data-processing. On one hand, the bridge health monitoring makes use of many sophisticated sensors and data collecting equipment; on the other hand, it ignores the reasonable processing and analysis of the data collected, leading to the failure to master the accurate bridge operation status and seriously constraining the security effect from the bridge health monitoring system on the bridge daily operation. In order to support the structural early-warning and evaluation systems, it’s desirable to develop a set of efficient, comprehensive and reliable data processing method, analyse the regularity of monitoring data and dig the information that can reflect the real working state of the bridge.Based on Xinguang Bridge health monitoring system, the method of data processing is systematically studied and the data processing software is developed. Specific working descriptions are as follows:(1) The research history of bridge health monitoring data processing method is presented first.(2) Research on the bridge health monitoring data pre-processing methods is implemented. The problems of missing data, abnormal data and noise interference appeared in Xinguang bridge health monitoring system are solved.(3) Research on temperature data processing methods is implemented. Based on actual measured temperature data, the atmospheric temperature distribution and component temperature field are obtained. Research on wind speed and direction data processing methods is implemented. Based on actual measured wind speed and direction data, the average and fluctuating wind characteristics of typhoon at the bridge location are obtained. Research on traffic flow data processing methods is implemented. Based on traffic-lane occupancy model, the profile of traffic flow distribution and traffic capacity of Xinguang Bridge are obtained. Research on stress data processing methods is implemented. Based on actual measured strain data, the statistical parameters of the component stress data are obtained. Research on modal parameter identification methods is implemented. Based on the stochastic subspace identification and the peak pick identification, the first ten order modal parameter of Xinguang Bridge are obtained. Research on tension identification method is implemented. Based on frequency method, the suspender force-frequency model of Xinguang Bridge is proposed and tensions of suspender are obtained.(4) Research on the relevance between temperature and stress, between temperature and suspender force, between humiture and bridge dynamic characteristics according to the data from Xinguang Bridge is implemented. Based on that, the method of separating the component temperature stress, constant stress and live stress is proposed. Based upon the method, temperature stress, constant stress and live stress of the component are obtained.(5) Based on the characteristics of Xinguang Bridge health monitoring data, a bridge health monitoring data processing software is developed on the platform of MATLAB GUI to process the bridge health monitoring data efficiently and reliably.Based on the work accomplished, following conclusions are obtained:(1) There is same difference in temperature longitudinally and vertically between the components of Xinguang Bridge. Therefore, in the analysis of temperature effect, not only the heating and cooling temperature of the whole bridge but also the temperature difference of the partial components should be considered. There exists a time-lag effect between the temperature of concrete members and atmospheric temperature.(2) There some differences between the longitudinal wind power spectrum of typhoon at the Xinguang Bridge location and the Simiu spectrum regularly recommended.(3) The maximum capacity of traffic flow of the single lane on the Xinguang Bridge is 1792 vehicles per hour. The actual measurement of the traffic-lane occupancy ratio shows that the traffic condition on Xinguang Bridge is good and the traffic congestion rarely occurs.(4) Temperature is the main factor that influences the variation of component stresses. And the relationship between stress and environmental temperature is almost linear. Hence, linear fitting can be used to separate stress to get temperature stress, constant stress and live stress of components.(5) The peak pick identification and the stochastic subspace identification respectively belong to frequency-domain and time-domain mode identification methods. In the identification of intensive modal structure parameters, the peak pick identification reflects in the relatively low accuracy and the phenomenon of “order leakage”. In comparison, the stochastic subspace identification directly uses real-time data to implement the modal parameter identification, which shows a relatively high accuracy.(6) The influence of temperature on Xinguang Bridge varies according to different model frequencies. The atmospheric humidity has a certain degree of influence on the modal frequency of Xinguang Bridge.(7) With the rising temperature, the whole Xinguang Bridge expands. The fact that the relative deformation between the steel arch rib and the bridge decking is greater than the temperature expansion of the cable has resulted in the increase of suspender force.