A Neural-network-based System for Bridge Health Monitoring

A Neural-network-based System for Bridge Health Monitoring

A bridge health monitoring system based on neural network technology is proposed in this research. Nowadays, accompanying with the aging of the existing bridges all over the world, how to effectively identify the health condition of the bridges has become an important issue. The method should offer a rapid and reliable result immediately after major strikes without using lots of labor and time. The demand of this health monitoring system grows rapidly and research on this topic has been widely discussed. Meanwhile, neural networks, commenced from artificial intelligence, have also shown their outstanding performance in complex problems. For this reason, a monitoring system using neural network is developed. As commonly known, the strong motion recording network of structures and bridges in Taiwan has offered an excellent database for health monitoring. Analytical result of different methods including transfer function, Autoregressive with Exogenous(ARX) model, and the proposed neural-network-based system are used to evaluate the efficiency in bridge health monitoring. The result has shown that the proposed neural-network-based system can be successfully used in bridge health monitoring after major earthquakes.