Research On The Health Monitoring System Of Hongfeng Lake Bridge And Its Key Technologies

World Bank issued a research report on China's freeway recently. According to the report, China's economic growth and poverty reduction over the last 15 years has made remarkable achievements in many areas. A most important area is infrastructure especially transport infrastructure. But 19 years ago, the freeway has not even one meter. From 1990 to 2005, China has build 41,000 km of freeway, ranking second in the world, next only to the United States. In the future, this figure will double to reach 85,000 km in which 40,000 km will be built, of which half of these concentrated in the western region. This means that in the western region also need a lot of bridges, tunnels. Meanwhile, many long-span bridge be built in the freeway have been in or are about to the safety maintenance period. Clearly, that the author engaged in Hongfeng Lake Bridge health monitoring system and the research of its key technologies is important to the western region of China in a practical sense, especially to Guizhou Province.After entering the 21st century, because the rapid development of China's freeway many scholars began to pay attention to the long-span bridge health monitoring and maintenance. At the same time, bridge project involves many subjects, and the structural health monitoring is complex, many of the pivotal theory and method still in the exploration. Based on Guizhou Hongfeng Lake Bridge health . monitoring system design and implementation, the paper has conducted fruitful research in the single-tower long span cable-stayed bridge health monitoring systems and methods. The main work to complete and innovation are as follows:Through detailed research and field study, decide to select the cable-stayed bridge to implement the health monitoring system. The analysis and design of Hongfeng Lake Bridge real-time health monitoring system has completed that consist of the cable force monitoring subsystem, the main beam vibration monitoring subsystem, stress monitoring and its temperature compensation subsystem. The achievement is a demonstration project for long-span bridge health monitoring system in Guizhou Province. FBG sensors and SDH (Synchronous Digital Hierarchy) communication technology were used to solve the problem how to transmit the bridge health monitoring system's real-time information in ditch canyon to the monitoring center.China Mobile Communication and Internet Short Message technology were used to solve the problem How to transmit the bridge health monitoring system's real-time information into web site. Then the users of related can use the Information in the Internet and diagnose the bridge's health.FFT (Fast Fourier transform) was used to process the stay cables' real-time signal from the acceleration sensor. To ensure the accuracy of cable force, the method that combination of peak picker, cepstrum and auto-correlation function was used to identify the cable natural frequency. By the comparison of this analysis method, have found the flaws inherent in the application of the peak picker method, and, the auto-correlation function and cepstrum should be used as a complementarity.Due to the problem of fund, it is impossible to have acceleration sensors installed on every cable. In order to get each cable's force, this paper presents a three-layer BP neural network model and trains the neural network with a number of measured data. By using this neural network model, the cable force of not installed acceleration sensor can be derived. The neural network model is certificated by using the measured data. By comparison with measured data and the derived data, the cable force forecasts from the neural network are credible.The aim of bridge health monitoring system is the structure damage identification. The key for damage identification is modal parameter identification. EMD is used to the bridge vibration signal to derive the IMF and the residue. And further RDT and HT are used to derive natural frequency and damping ratio. Proved by the actual measured data, HHT method is a good for large-scale bridge modal parameter identification.