A Hybrid Data-Driven-Agent-Based Modelling Framework for Water Distribution Systems Contamination Response during COVID-19

Source identification of water distribution system contamination based on simulated annealing-particle swarm optimization algorithm

Ensuring the safety of water supplies is critical for urban areas requires rapid response when water quality anomalies are detected in the pipeline network. Prompt action is essential to prevent widespread contamination, protect public health, and mitigate potential social unrest. The particle swarm optimization (PSO) algorithm has faced challenges for contamination source identification (CSI) in water distribution systems (WDS), primarily due to its susceptibility to locally optimal solutions. Addressing this issue is critical to quickly and accurately identify contamination sources. Therefore, this research integrates the Metropolis criterion from the simulated annealing (SA) algorithm into a SA-PSO algorithm, to overcome the limitations of PSO. This study conducts contamination localization experiments using SA-PSO, with the publicly available NET-3 pipeline network as the case to generate sudden contamination events. By collecting pollutant concentration data from predefined monitoring points over time through simulation, a simulation-optimization inverse location model is constructed to fit the pollutant concentrations at each monitoring point. The results of the case study show that SA-PSO outperforms PSO in both speed and accuracy in solving the CSI problem, and the findings provide an efficient and effective contamination localization tool for urban water supply management.