We are developing a wireless data gathering system from household terminals for continuous monitoring and active control of city lifelines for a disaster situation. The gathered information creates a database to control lifeline systems and to make an effective restoration plan by GIS. The lifeline includes a large network of water supply. This paper proposes a method for reducing the network size into a feasible size by linear approximation and a pattern matching method for damage identification. The evaluation of the method through implementation shows the execution time, the feasibility of producing a prediction database and the accuracy.
展开▼