Advancing Robust Multi-Objective Optimisation Applied to Complex Model-Based Water-Related Problems  book cover
1st Edition

Advancing Robust Multi-Objective Optimisation Applied to Complex Model-Based Water-Related Problems

ISBN 9780367460433
Published March 3, 2020 by CRC Press
164 Pages

SAVE ~ $15.99
was $79.95
USD $63.96

Prices & shipping based on shipping country


Book Description

The exercise of solving engineering problems that require optimisation procedures can be seriously affected by uncertain variables, resulting in potential underperforming solutions. Although this is a well-known problem, important knowledge gaps are still to be addressed. For example, concepts of robustness largely differ from study to study, robust solutions are generally provided with limited information about their uncertainty, and robust optimisation is difficult to apply as it is a computationally demanding task.

The proposed research aims to address the mentioned challenges and focuses on robust optimisation of multiple objectives and multiple sources of probabilistically described uncertainty. This is done by the development of the Robust Optimisation and Probabilistic Analysis of Robustness algorithm (ROPAR), which integrates widely accepted robustness metrics into a single flexible framework. In this thesis, ROPAR is not only tested in benchmark functions, but also in engineering problems related to the water sector, in particular the design of urban drainage and water distribution systems.

ROPAR allows for employing practically any existing multi-objective optimisation algorithm as its internal optimisation engine, which enables its applicability to other problems as well. Additionally, ROPAR can be straightforwardly parallelized, allowing for fast availability of results.

Table of Contents

1 Introduction
1.1 Background
1.2 Motivation
1.3 Research questions
1.4 Objectives
1.5 Innovation, practical value and social relevance
1.6 Thesis structure

2 Literature review
2.1 Evolutionary algorithms
2.2 Main types of multi-objective optimisation (MOO) algorithms
2.3 Robust multi-objective optimisation (RMOO)
2.4 Conclusions

3 Methodology
3.1 Methodological framework
3.2 Identification and characterization of uncertainty sources
3.3 General formulation of the optimisation problem
3.4 Mathematical definitions of robustness metrics
3.5 Selection of MOO algorithm
3.6 Formulation of the proposed RMOO algorithm
3.7 Reliability of satisfying constraints
3.8 Dealing with more than two objective functions
3.9 Experimental plan
3.10 Exemplifying ROPAR
3.11 Analysis of computational complexity

4 Robust optimisation of a simple storm drainage system
4.1 Problem statement
4.2 Experimental setup
4.3 Results and discussion
4.4 Conclusions

5 Robust optimisation of two larger storm drainage systems
5.1 Problem statement
5.2 Experimental setup
5.3 Case studies
5.4 Results and discussion
5.5 Conclusions

6 Robust optimisation of a storm drainage system: more objectives and sources of uncertainty
6.1 Problem statement
6.2 Experimental setup
6.3 Case study
6.4 Results and discussion
6.5 Conclusions

7 Robust optimisation of water quality in distribution systems
7.1 Water age minimisation problem and its deterministic solution
7.2 Using ROPAR with the new MOO, two objective functions, twenty four sources of uncertainty
7.3 Conclusions

8 Conclusions and recommendations
8.1 Summary
8.2 Conclusions
8.3 Limitation of this study
8.4 Recommendations

View More



Oscar Osvaldo Marquez Calvo has a Bachelor Degree (1990) and a Master Degree in Computer Sciences (1992). He also has a Master in Manufacturing Systems (1995). Since then, he worked in the field of Information Technology in technical operations and application development as well. In 2011, he started working in the area of Hydroinformatics. And he finished in 2020 his PhD in a joined program between IHE Delft Institute for Water Education (IHE Delft) and Delft University of Technology (TU Delft), Netherlands. His areas of interest are robust optimization, computer modelling and artificial intelligence.