Department of Civil Engineering, Tianjin University, Tianjin 300072, China
Abstract:
As one of the most important urban lifeline systems, a water distribution system can be damaged under a strong earthquake, and the damage cannot easily be located, especially immediately after the event. This often causes tremendous difficulties to post-earthquake emergency response and recovery activities. This paper proposes a methodology to locate seismic damage to a water distribution system by monitoring watcr head online at some nodes in the water distribution system. An artificial neural network-based inverse analysis method is developed to estimate the water head variations at all nodes that are not monitored based on the water head variations at the nodes that are monitored. The methodology provides a quick, effective, and practical way to locate seismic damage to a water distribution system.