Redesigning Water Distribution Networks Using a Guided Evolutionary Approach

Abstract

The paper describes and discusses the application of a guided evolutionary approach for the optimization of water distribution networks. Initially developed for the Battle of the Water Networks II (BWN-II), the approach was adapted and improved for competing in the Battle of Background Leakage Assessment for Water Networks (BBLAWN). The proposed evolutionary approach operates in an exclusively discrete solution space and is intended to require as little engineering judgment and time as possible while attaining acceptable and informative results that are useful for decision making. Its main features are custom crossover and mutation operators specific for water distribution network optimization tasks. A simple postprocessing greedy algorithm to locally refine pipe replacements is introduced as a means of complementing the evolutionary approach.