This paper describes the development of an adaptive locally constrained genetic algorithm (ALCO-GA) and its application to the problem of least cost water distribution network design. Genetic algorithms have been used widely for the optimisation of both theoretical and real-world nonlinear optimisation problems, including water system design and maintenance problems. In this work we propose a heuristic-based approach to the mutation of chromosomes with the algorithm employing an adaptive mutation operator which utilises hydraulic head information and an elementary heuristic to increase the efficiency of the algorithm's search into the feasible solution space. In almost all test instances ALCO-GA displays faster convergence and reaches the feasible solution space faster than the standard genetic algorithm. ALCO-GA also achieves high optimality when compared to solutions from the literature and often obtains better solutions than the standard genetic algorithm.
Adaptive locally constrained genetic algorithm for least-cost water distribution network design
Matthew B. Johns, Edward Keedwell, Dragan Savic; Adaptive locally constrained genetic algorithm for least-cost water distribution network design. Journal of Hydroinformatics 1 March 2014; 16 (2): 288–301. doi: https://doi.org/10.2166/hydro.2013.218
Download citation file: