Back to search

National Technical University of Athens (NTUA)

Surrogate-based optimization methods for coastal aquifer management

Abstract

dc:description

Iterative methods, such as sensitivity analysis, uncertainty analysis or optimization, require the use of high-fidelity physics-based models for applying them effectively in complex physical systems. The coupling of such methods with detained numerical models is often impractical due to the excessive computational cost. The present thesis deals with a similar problem arising from the use of high-fidelity models of coastal aquifer flow in pumping optimization of coastal aquifers where the standard simulation-optimization routines are hindered by the increased computational burden. The thesis focuses on the use of surrogate models or metamodels, which is a widely acknowledged and successful approach in the relevant literature, for reducing the computational effort during optimization. The main aim of this work is the development of effective and efficient surrogate-based optimization (SBO) frameworks for coastal aquifer management. The existing literature of SBO is extensively discussed for the broader field of engineering optimization and the relevant research in water resources management. Emphasis is given on the development of SBO methods for pumping optimization of coastal aquifers while an exhaustive comparison among the various approaches is conducted for problems of different dimensionality and available computational budget with the computationally demanding numerical model of coastal aquifer flow. It is noted that the effectiveness of the proposed methods is tested based on multiple independent optimization trials. The following SBO schemes are developed: (a) A surrogate-assisted evolutionary framework using a pure exploitation infill strategy. (b) A surrogate-assisted evolutionary framework based on a multiple surrogate approach (either selecting the best surrogate model or forming an optimal weighted ensemble of surrogate models) and a pure exploitation infill strategy. (c) An adaptive-recursive optimization framework which balances exploration with exploitation using the metamodels. (d) Multi- fidelity optimization based on seawater intrusion models of different fidelities and co-Kriging models. The proposed SBO methods delivered reliable solutions in the region of the global optimum using a small number of simulations with the computationally intensive variable density and solute transport model. The presence of multiple nonlinear constraints in pumping optimization of coastal aquifers was successfully treated with the use of radial basis functions as metamodels and the proposed optimization frameworks. The performance of the SBO algorithms that are developed in this thesis was comparable to published SBO methods which have been widely used in the relevant literature. The novel multi-fidelity optimization scheme based on co-Kriging shown great promise for its application in pumping optimization problems of coastal aquifers of extremely high computational cost.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2021

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Christelis, Vasileios
  • Χριστέλης, Βασίλειος

Subjects

dc:subject × 14

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/49304

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Christelis, Vasileios; Χριστέλης, Βασίλειος. Surrogate-based optimization methods for coastal aquifer management. National Technical University of Athens (NTUA), 2021. http://hdl.handle.net/10442/hedi/49304