Comparing Optimization Techniques with an Engineering Judgment Approach to WDN Design

Abstract

The design of water distribution networks (WDNs) is a complex nonlinear problem that has been solved using optimization techniques attempting to find near-globally optimal solutions. This paper demonstrates that relying on engineering judgment, with very limited use of optimization, can generate reasonable quality approximations of the Pareto front of multiobjective WDN design problems without intensive computational requirements. First the results of the implementation of a previous implementation of the engineering judgment approach to the Battle of Background Leakage Assessment for Water Networks (BBLAWN) design competition is summarized. Additionally, a simplified version of the BBLAWN problem is solved using both the engineering judgment approach and a multiobjective optimization algorithm and the corresponding Pareto fronts are compared. Relative to other BBLAWN solutions, the engineering judgment solution produced a high-quality result with a very small computational budget and was able to identify a more realistic number of changes in the WDN (∼10% of the network) compared with BBLAWN teams relying on global optimization algorithms.