Bridge Health Monitoring System Based On Damage Analysis

During the last36years, the reform and open policy has brought us Chinesepeople a rich and colorful life. With the rapid development of economy, bridgesplay more and more important role in the economic development of our country.Due to the material aging, excessive use, long use fixed number of year, thedamage of the bridge is more and more serious, and even many bridge collapseaccidents have occurred. These brought serious threat to the transportation of ourcountry and people’s life and property safety. In order to prevent the malignantaccident, Health monitoring method which includes real-time structural safetyparameters monitoring of the bridge has become a very important topic recently.In this paper, we design the overall architecture of bridge health monitoringsystem based on the study of actual bridge information and existing research.This paper selects the Fiber Bragg Grating sensor as the primary data collectionunit which collects the data of the bridge sensitive parameters. We transfer thecollecting data back to the bridge monitoring center through the TCP/IP and storethe data by using the SQL2000. And last display the collecting information in themonitoring center computers by using Force-control7.0software.Except the data monitoring system, this paper set up the data processingcenter in the monitoring center. Due to the excellent ability of pattern recognitionof neural network, this paper selects neural network use the neural networkalgorithm to detect the bridge damage by analysis the structure information of thebridge. We take the damage location and damage degree as the input of the neuralnetwork, and take the difference between the health structure and the one withdamage as the output of the neural network. In the paper, we use the ANSYSsoftware modeling and analysis the bridge. And then train and study the sampledata to realize the identification of the structure parameters of the bridge by using the genetic algorithm and neural network algorithm. In the end of this paper,verify the correctness and practicability of the neural network and geneticalgorithm through the simulation analysis of an actual bridge. As is verified, themonitoring method and damage identification method discussed in this paperhave the very good application value for the reinforced concrete continuousgirder bridge. And this kind of the monitoring method also have certain referencein other areas such as tunnel, building, DAMS monitoring.