Prof. Kyriakos Giannakoglou received his B.S. degree in mechanical engineering in 1982 and his Ph.D. degree in Computational Fluid Dynamics in 1987, both from the National Technical University of Athens (NTUA), Greece. He is Professor with the Lab. of Thermal Turbomachines and head of the Parallel CFD and the Optimization Unit of the School of Mechanical Engineering of NTUA. His research interests include development of CFD methods for turbomachinery applications as well as external aerodynamics (including car applications), development of inverse design and optimization algorithms based on evolutionary algorithms and deterministic (adjoint) methods and/or neural networks and parallelization of the corresponding software (including Cluster & Grid Computing and GPUs). His research group has developed and brought to market the generic optimization software EASY (Evolutionary Algorithm SYstem). Major part of the research carried out by his group is being funded by European industries. Since May 2007, he is the Chairman of the ERCOFTAC (European Research Community of Flow Turbulence and Combustion) Special Interest Group SIG34 on Design Optimization. He has authored ~80 Journal papers, ~150 Conference papers and several chapters in books. 

Summary
This lecture presents the formulation, validation and application of continuous adjoint in aero/hydrodynamic shape optimization and robust design. Regarding shape optimization, emphasis is laid on some particular features, the omission of which damages the accuracy of the computed gradient. The first one is the handling of turbulence models in continuous adjoint. Unlike discrete adjoint, the majority of continuous adjoint methods rely upon the assumption of frozen turbulence which might become quite inaccurate and, in some cases, might lead to wrongly signed sensitivity derivatives. The differentiation of the most frequently used turbulence models and the corresponding benefits are discussed and demonstrated. A second fundamental issue in continuous adjoint is related to the inclusion of grid sensitivities into the adjoint formulation. Regarding robust optimization, this lecture focuses on the Method of Moments (MoM) that requires higher derivatives of the Quantity of Interest (QoI). Apart from the computation of the higher derivatives at the minimum computational cost, an algorithm (in the first-order MoM) that avoids computing second-order derivatives by computing matrix-vector products instead, using a combination of adjoint and direct differentiation will be demonstrated. This makes the cost per gradient computation independent of the numbers of both the design and uncertain variables. Among other, applications for the automotive industry will be presented.

Maria João Alves received a degree in Computers Engineering and a PhD in Management (Operational Research) from the University of Coimbra. She is assistant professor at the Faculty of Economics of the University of Coimbra and co-coordinator of the PhD in Management - Decision Aiding Science. Her main research interests include multiobjective optimization, evolutionary algorithms, bilevel optimization and applications, in particular for the design of electricity retail pricing. She is co-author of the book Multiobjective Linear and Integer Programming published by Springer in April 2016.

Summary
Bilevel optimization enables to model hierarchical non-cooperative decision processes in which there are two decisions makers (the leader and the follower) who control different sets of variables and have their own objective functions. Decisions are made sequentially, as the leader makes his decisions first by selecting values for his variables. The follower then reacts by optimizing his objective function on the feasible solutions restricted by the leader. Therefore, the leader must anticipate the follower's reaction because it affects the leader's objective value. Sequential decision-making processes that can be modeled by bilevel optimization problems arise in many aspects of resource planning, management and policy-making, namely in the design of pricing policies.
In this talk, we will present and illustrate the main concepts of bilevel optimization and describe an application in the electricity retail market . A bilevel optimization model with multiple objective functions at the lower level has been developed to study the interaction between an electricity retailer and consumers engaging in demand response. The retailer (leader) establishes dynamic time-of-use tariffs to maximize profits. The consumer (follower) responds by selecting, under that price setting, an appliance scheduling aiming to minimize the electricity bill and minimize the discomfort associated with operating appliances outside habitual patterns, to make the most of time-differentiated prices. The existence of multiple objective functions at the lower level problem adds further challenges to the decision process due to the uncertainty for the leader related to the follower's reaction, i.e., the choice of a nondominated solution trading-off cost and discomfort. A hybrid approach to compute different types of solutions is presented, which integrates a genetic algorithm for the upper level search with an exact solver to determine optimal solutions to scalarization problems at the lower level.

José António Covas is full professor at the Department of Polymer Engineering, University of Minho, Portugal, since 1998. His research interests include polymer extrusion, additive manufacturing, preparation of nanocomposites and in-process characterization. He has authored 200 papers in international journals with peer reviewing, books and book chapters, as well as 11 patents. He is Editor-in-Chief of the International Polymer Processing Journal and member of the Editorial Board of three other scientific journals. He is Vice-President of PIEP (Innovation Pole in Polymer Engineering), a private non-profit R&D organization. He is an expert/reviewer for a number of national funding agencies. He has intense cooperation with industry via projects aiming at product or materials development and process monitoring and optimization.

Summary
Polymers are an important class of materials with a wide application in the packaging, building and construction, health, energy, transportation, agriculture, electrical and electronic, household, leisure and sports industries. It is estimated that more than 60000 companies in Europe, most of them SMEs, work with polymers. Major polymer processing technologies, such as extrusion, injection moulding and blow moulding operate on the common principle of melting the material (usually supplied in pellet form), shape it and cool it down. Thus, each of these techniques entails the flow and heat transfer along an intricate geometry of a material with a complex rheological behaviour. Research on process modelling has generated a better understanding of the phenomena developing in these techniques. However, modelling by itself cannot be used to efficiently to design better equipment, or to set the most adequate operating conditions. Indeed, solving the inverse problem, i.e., solving the governing equations of the process in order to the geometrical variables and/or operating parameters, is generally mathematically ill-posed. An alternative approach consists in linking global optimization methods and process modelling. Yet, the application of multiobjective optimization to polymer processing is not straightforward due to the nonlinear interactions between the search variables, the multimodality of the search space, the usual high number and conflicting character of the objectives and the need to meet convergence and diversity requirements for the solutions. Several practical polymer processing case studies will be presented and the challenges of the approach will be discussed.

Massimiliano Vasile is a Professor of Space Systems Engineering and Director of the Aerospace Centre of Excellence in the Department of Mechanical & Aerospace Engineering at the University of Strathclyde. Prior to this, from 2005 to 2010, he was Head of Research for the Space Advanced Research Team at the University of Glasgow. Before starting his academic career in 2004, he was the first member of the ESA Advanced Concepts Team and initiator of the ACT research streams on global trajectory optimization, mission analysis and biomimicry. He received his M.S. in 1996 and Ph.D. in 2001 from Politecnico di Milano. He sits on the IAF Astrodynamics and Space Power committees, the IEEE committee on Emerging Technologies in Computational Intelligence, and the UN Space Mission Planning Advisory Group. His research interests include Astrodynamics, Space Traffic Management, Computational Intelligence and Optimization Under Uncertainty exploring the limits of computer science at solving highly complex problems in science and engineering. Asteroid 2002 PX33 "Maxvasile" was named in his honour in recognition of Prof Vasile's contributions to the development of innovative techniques for the design and optimization of space trajectories and his work on asteroid manipulation. He is a Senior Member of AIAA.

Summary
This talk will present some recent advances on the solution of multi-objective optimal control problems under epistemic uncertainty. The talk will start with the presentation of a direct transcription method for the solution of constrained multi-objective optimal control problems with some theoretical results on local convergence and optimality of the solution. It will then introduce the idea of epistemic uncertainty and how this can be incorporated in the solution of optimal control problems. In particular, in this talk, epistemic uncertainty is modelled with p-boxes and will affect both the value of the cost function and the satisfaction of the terminal constraints. A couple of illustrative examples will conclude the talk.

Carlos M. Fonseca is an Associate Professor at the Department of Informatics Engineering of the University of Coimbra, Portugal, and a member of the Evolutionary and Complex Systems (ECOS) group of the Centre for Informatics and Systems of the University of Coimbra (CISUC). He graduated in Electronic and Telecommunications Engineering from the University of Aveiro, Portugal, in 1991, and obtained a Ph.D. in Automatic Control and Systems Engineering from the University of Sheffield, U.K., in 1996. His research has been devoted mainly to evolutionary computation and multi-objective optimization, with a focus on computationally efficient approaches to preference articulation and experimental performance evaluation in evolutionary multi-objective optimization. He is the Scientific Representative of the Grant Holder of COST Action CA15140 - Improving Applicability of Nature-Inspired Optimisation by Joining Theory and Practice (ImAppNIO), and the leader of a Working Group on Software in that Action. He has served as General, Technical or Track co-Chair of several major international conferences on evolutionary computation, and is a member of the Evolutionary Multi-Criterion Optimization and of the Parallel Problem Solving from Nature Steering Committees.

Summary
Selection plays a double role in evolutionary algorithms. The selection of solutions from the current solution set (or population) to produce new candidate (offspring) solutions through variation is known as parental selection , whereas the selection of solutions to discard in order to make room for new solutions in the population is usually called environmental selection . Intimately related to selection is the concept of solution fitness, which is typically related to the objective function(s) of the problem at hand. In practice, both types of selection must be implemented, although different evolutionary algorithms often emphasize either one or the other. In particular, it is common for only one type of selection to depend on fitness. Selection also has the double purpose of steering the search towards more promising regions of the search space by favouring the best solutions available ( exploitation ) while maintaining a sufficient level of diversity in order to be able to escape local optima and/or find multiple good solutions ( exploration ). Over the years, many different approaches to selection in evolutionary algorithms have been proposed in the literature, with the balance between exploration and exploitation gaining heightened importance in the context of multiobjective optimization. However, parental and environmental selection have continued to be seen as different operators, and to be implemented separately. In this talk, selection methods and fitness assignment strategies are reviewed and discussed from the unifying perspective of portfolio optimization, where the fitness of a solution is interpreted as an investment in that solution, and solution diversity emerges naturally from the need to balance return and risk in the portfolio. In addition, parental and environmental selection can be seamlessly integrated in the portfolio optimization formulation. Application examples illustrate the main aspects of the approach.