Multi-objective Optimal Scheduling of Valves and Hydrants for Sudden Drinking Water Pollution Incident

Abstract

In the last decades, water pollution incidents have occurred frequently, causing severe significant economic losses, and negative social influence. How to establish and improve emergency disposal mechanism for water pollution incident is an important issue, which has become a foremost concern in the world. In this paper, we first make a theoretical analysis on the optimal scheduling of valves and hydrants and prove that it is NP-Complete, then a multi-objective optimization model for contaminant response is established and two conflicting objectives are explored: (1) minimization of the volume of contaminated water exposure to the public, (2) minimization of the costs of operations on hydrants and valves which needed for isolation and flushing of the contaminant in the water distribution network. Finally, a customized multi-objective non-dominated sorted genetic algorithm-II (NSGA-II) linking to EPANET simulation is utilized to trade off the two optimization objectives. A medium size of water distribution network is employed for demonstrating the validity of proposed model and methodology.

Xuesong Yan received him B.E. degree in Computer Science and Technology in 2000 and M.E. degree in Computer Application from China University of Geosciences in 2003. He received his Ph.D. degree in Computer Software and Theory from Wuhan University in 2006. He is currently with School of Computer Science, China University of Geosciences, Wuhan, China and was as a visiting scholar with Department of Computer Science, University of Central Arkansas, Conway, USA. His research interests include evolutionary computation, data mining and computer application.