Answers are grounded in this paper and cite their sources.
The optimal design of water distribution networks involves minimizing the cost of the network subject to the constraints of satisfying the pressure requirements at all the nodes. In genetic algorithm (GA)-based optimal design, such constrained problems are generally converted to unconstrained ones using penalty methods. The penalty is applied for not meeting the pressure constraints. Researchers observed previously that the optimal solutions lie on the boundary of feasible and infeasible solutions. They suggested self-organizing penalty based on pressure heads for increasing the computational efficiency and obtaining better solutions. Usually, a flow-dependent analysis is carried out to obtain pressure deficiency at different nodes with the assumption that all nodal demands are satisfied. The assumption of demand satisfaction at all nodes for infeasible solutions results in higher pressure deficiency at different nodes, resulting in higher penalty costs than actually required. This may delay the convergence. Herein, a novel flow-based penalty approach is suggested in which the penalty is obtained based on the shortfalls in the pressures and flows at different nodes using pressure-dependent analysis. The methodology is illustrated with results from benchmark networks. The modified GA method is found to be converging faster as compared to other methods of penalty application.
No takes yet. Share an insight, caveat, or question.
Sayyed et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: