Optimum Sanitary Sewer Network Design Using Shuffled Gray Wolf Optimizer

Abstract

Sanitary sewer networks are among the most important and most costly infrastructures in the water resources management field. However, the optimum design of these networks is cumbersome and sophisticated; due to these systems’ sensitive and costly construction, the application of optimization algorithms is essential. Focusing on solving this nonlinear optimization problem with continuous and discrete constraints, the shuffled gray wolf optimizer (SGWO), which is a hybrid algorithm inspired by shuffled complex evolution (SCE) and gray wolf optimizer (GWO), was used to solve two hypothetical networks and one real sewer network with different scales and geometries and was compared to four well-known metaheuristic optimization algorithms. The results showed that the population division in the SGWO algorithm not only improved the results obtained by the GWO and other algorithms but also the costs were lower than those achieved using other famous optimization algorithms. This happened while the SGWO algorithm used a considerably smaller number of function evaluations. Moreover, this algorithm exhibited low standard deviation and average objective function value in 20 independent runs, which shows this algorithm’s reliability.