Improving the efficiency of multi-objective evolutionary algorithms through decomposition: An application to water distribution network design. (July 2015)

• Record Type: Journal Article
• Authors: Zheng, Feifei; Simpson, Angus; Zecchin, Aaron
• Is Part Of: Environmental modelling & software. Volume 69(2015:Jul.)
• Page Start: 240
• Page End: 252
• Publication Date: 2015-07
• Subjects: Graph decomposition -- Water distribution network -- Differential evolution algorithm -- Multi-objective optimization; Environmental monitoring -- Computer programs -- Periodicals; Ecology -- Computer simulation -- Periodicals; Digital computer simulation -- Periodicals; Computer software -- Periodicals; Environmental Monitoring -- Periodicals; Computer Simulation -- Periodicals; Environnement -- Surveillance -- Logiciels -- Périodiques; Écologie -- Simulation, Méthodes de -- Périodiques; Simulation par ordinateur -- Périodiques; Logiciels -- Périodiques; Computer software; Digital computer simulation; Ecology -- Computer simulation; Environmental monitoring -- Computer programs; Periodicals; Electronic journals; 363.70015118
• DOI: 10.1016/j.envsoft.2014.08.022
• Languages: English
• ISSNs: 1364-8152

Abstract

Abstract: Evolutionary algorithms (EAs) have been widely used in handling various water resource optimization problems in recent years. However, it is still challenging for EAs to identify near-optimal solutions for realistic problems within the available computational budgets. This paper introduces a novel multi-objective optimization method to improve the efficiency of a typically difficult water resource problem: water distribution network (WDN) design. In the proposed approach, a WDN is decomposed into different sub-networks using decomposition techniques. EAs optimize these sub-networks individually, generating Pareto fronts for each sub-network with great efficiency. A propagation method is proposed to evolve Pareto fronts of the sub-networks towards the Pareto front for the full network while eliminating the need to hydraulically simulate the intact network itself. Results from two complex realistic WDNs show that the proposed approach is able to find better fronts than conventional full-search algorithms (optimize the entire network without decomposition) with dramatically improved efficiency. Highlights: The graph decomposition dramatically improves the optimization efficiency. The propagation method effectively evolves sub-network fronts to the full network front. The optimization strategy is demonstrated using real-world WDNs with high complexities. Provide an efficient decision-making tool for the water network optimization.