Self-Adaptive Models for Water Distribution System Design Using Single-/Multi-Objective Optimization Approaches

Choi Choi,Kim Kim

doi:10.3390/w11061293

Abstract

This study compares the performance of self-adaptive optimization approaches in efficient water distribution systems (WDS) design and presents a guide for the selection of the appropriate method employing optimization utilizing the characteristic of each technique formulation. To this end, this study performs three types of analyses. First, the sensitivity analysis of each self-adaptive approach is conducted on single/multi-objective mathematical benchmark problems with various problem types (e.g., using solution shape or many local optimal solutions). Second, based on the applications and results of the mathematical problem, the performance of the algorithm is verified in the WDS design problem considering the minimum cost and the maximum system resilience under the single/multi-objective optimization framework. Third, the characteristics of search operators in the self-adaptive approach are compared according to the presence or absence of additional parameters and operators. Moreover, various performance indices are employed to compare the quantitative evaluation of each algorithm. Each algorithm is found to exhibit different characteristics depending on the problem scale and solution type. These results are expected to benefit future research in the formulation of new approaches and developments. Hence, this study provides rigorous testing of the performance of newly proposed algorithms in a highly simplified manner.

Highlights

The water distribution system (WDS) is a system of engineered hydrologic and hydraulic components that enables the collection, transmission, treatment, storage, and distribution of water from the source to demand locations, while satisfying the nodal pressure and pipe velocity
Based on the applications and results of the mathematical problem, the performance of the algorithm is verified in the WDS design, considering the minimum cost and maximum system resilience under the single/multi-objective optimization framework
The characteristics of the search operators of the self-adaptive approach are compared according to the presence or absence of additional parameters and operators

Summary

Introduction

The water distribution system (WDS) is a system of engineered hydrologic and hydraulic components that enables the collection, transmission, treatment, storage, and distribution of water from the source to demand locations, while satisfying the nodal pressure and pipe velocity. Various design techniques have been developed and applied for effective WDS design in the past. Traditional approaches differed in the quality of their solutions. They did not provide stable designs and could not guarantee optimality. To overcome this kind of drawback, the calculation-based approaches (i.e., linear programming, non-linear programming, and dynamic programming) were applied for the optimal design of WDSs [1,2,3,4,5,6,7,8,9,10]. The application of these techniques in real-world problems renders the WDS models too complex to be handled by conventional methods. More efficient techniques to solve this kind of design problem must be investigated

Methods

Results

Conclusion