Application of the NSGA-II multi-objective algorithm in the optimal design of drinking water distribution networks. Case: Huancavelica City, Peru

Tunque-Dueñas T.; Ricra-Dueñas F.; Ayala I.; Contreras-Lopez E.; Portuguez-Maurtua M. · Tecnologia y Ciencias del Agua (2025) · DOI: 10.24850/j-tyca-2025-02-04

Resumen

In recent times, several multi-objective genetic algorithms and their application in optimization of drinking water distribution networks have been developed, of which NSGA-II has shown the strongest performance. This research shows the application of NSGA-II in the optimal design of drinking water distribution networks considering cost (IC) and hydraulic reliability (IR) as objective functions. The research was carried out in response to a real problem related to water supply in the city of Huancavelica. Using the information obtained from EPS EMAPA Huancavelica S.A., and the Python programming language with the Epanet Toolkit, NSGA-II is validated by applying it to the design of the Hanoi network. Once validated, the Huancavelica network is analyzed, which has a IC of 0.31 equivalent to USD 140 099.89 and an IR of 0.25, and an optimal network design is obtained, which has a IC of 0.24 with a value of USD 117 590.12 and an IR of 0.23, which allows appreciating a difference in the IC of USD 22 509.77 and a reduction of the IR, which makes it a much more reliable network that simultaneously satisfies the minimum pressure restrictions in all the nodes, in addition to guaranteeing a capacity to withstand failure conditions during its operation. It is determined that NSGA-II is favorable for the optimal design of drinking water networks considering two objective functions of cost and hydraulic reliability.

Instituciones: Universidad Nacional Mayor de San Marcos; Universidad Nacional Agraria La Molina; Universidad Nacional de Huancavelica