Title: An ensemble-surrogate assisted cooperative particle swarm optimisation algorithm for water contamination source identification

Authors: Jinyu Gong; Xuesong Yan; Chengyu Hu

Addresses: School of Computer Science, China University of Geosciences, Wuhan, China ' School of Computer Science, China University of Geosciences, Wuhan, China ' School of Computer Science, China University of Geosciences, Wuhan, China

Abstract: The safety of water is vital to residents' health. The water supply network system could be accidentally or intentionally infringe easily, leading to the water contamination. Sensor networks can obtain useful observations to identify contamination events by detecting pollution in water supply networks, which can help to quickly identify the location and other related characteristics of contamination source. This paper proposes an ensemble-surrogate assisted particle swarm optimisation algorithm to solve water contamination source identification problem. Firstly, a modified collaborative particle swarm optimisation algorithm is proposed according to the characteristics of decision variables, which divides the three components reasonably, and adopts different evolutionary strategies to different components. Secondly, this paper uses ensemble-surrogate to solve computing expensive problem. An offline-model management is proposed to use less EPANET simulation. At last, experimental results show that the proposed algorithm is effective and efficient.

Keywords: water contamination source identification; expensive optimisation; surrogate-assisted evolutionary algorithm.

DOI: 10.1504/IJBIC.2022.123129

International Journal of Bio-Inspired Computation, 2022 Vol.19 No.3, pp.169 - 177

Received: 23 Oct 2021
Accepted: 23 Dec 2021
Published online: 30 May 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article