A two-stage post-disaster response and restoration method for the water distribution system

This paper proposed a two-stage restoration strategy based on China's post-earthquake repair experience. The first stage was the emergency stage. In order to efficiently calculate the pipes needed to be restored before the results of the biological optimization algorithm, this paper proposed a combination of the Dijkstra algorithm and the greedy algorithm. Dijkstra algorithm was used to calculate the water path from water source to hospitals and firefighting nodes. The water supply pressure of hospitals and firefighting nodes was guaranteed by no leakage of the water path. The restoration order of the remaining pipes in the first stage was determined by the greedy algorithm, and the second stage of the recovery strategy was determined by the particle swarm algorithm. The goal was to minimize the total amount of losses, and the pipes to be restored are assigned to three crews. In the simulation, the isolation, replacement, and reopening of one broken pipe were performed by the same crew. The entire simulation process was done by calling the EPANET program in Matlab. In the end, the total water loss rate during the entire restoration process was around 2.64%.