Sewer network design layout optimisation using ant colony algorithms
CITATION: De Villiers, N. B., Van Rooyen, G. C. & Middendorf, M. 2018. Sewer network design layout optimisation using ant colony algorithms. Journal of the South African Institution of Civil Engineering, 60(3):2-15, doi:10.17159/2309-8775/2018/v60n3a1.
The original publication is available at http://www.scielo.org.za
The optimal design of sewer networks typically comprises two sub-problems. The first is to determine an optimal layout of the network elements, and the second to optimally design the network components. In this article the focus is on the optimisation of gravity sewer network layouts, which requires simultaneous optimisation of hydraulic design. The layout is optimised using ant colony algorithms with four proposed node and edge-based selection strategies, while a heuristic optimisation algorithm is used for the hydraulic optimisation. The resulting simultaneous optimisation algorithm is shown to perform very well. The selection strategies are shown to be effective, but no clear best strategy is identified, as the performance of the layout algorithms is shown to depend heavily on characteristics of the network under consideration. However, some strategies are shown to perform inconsistently and worse than others on average.