Title Promoter Affiliations Abstract "Time parallel time integration methods for multiscale systems, with application in uncertainty quantification and optimization" "Giovanni Samaey" "Numerical Analysis and Applied Mathematics (NUMA), NUMA, Numerical Analysis and Applied Mathematics Section" "Time parallel methods iteratively refine a cheap but inaccurate time integration by computing a sequence of corrections on different portions of the time domain in parallel. Such methods promise extra opportunities for parallellisation on exascale high performance computing architectures. In this PhD, we will study time parallel time integration methods for multiscale systems and apply them in the context of uncertainty quantification and optimization, integrating the time parallellisation iterations within the optimization loop." "Regional climate modelling for wind farm optimization and control" "Nicole Van Lipzig" "Division of Geography and Tourism, Applied Mechanics and Energy Conversion Section" "The research concentrates on improving existing wind farm parameterization in the regional climate model COSMO-CLM, allowing for wind farm optimization and control. Calibration of model parameters will be done based on results from large eddy model simulation available from TFSO. COSMO-CLM is a non-hydrostatic mesoscale model and in this project, it will be used to study the interaction between large wind farm clusters and mesoscale weather systems like gravity waves, sea breezes, coastal fronts, convective and wind structures. The project will assess the effect of mesoscale weather systems on wind farm power extraction and study the potential for optimization and wind farm control. Moreover, it will assess the interaction between neighboring wind farms and get a better insight in optimal set point for wind farm operation taking into account the yield from the entire cluster instead of focusing on one farm." "Adaptive model reduction strategies for fast prediction and optimization of the vibro-acoustic performance of dynamical systems with damping" "Hui Zheng, Wim Desmet" "Production Engineering, Machine Design and Automation (PMA) Section, Mecha(tro)nic System Dynamics" "Over the past years, the dynamic and acoustic qualities of products have become an increasingly important design criterion in many industrial sectors. Moreover, in order to improve the cost-efficiency of products, lightweight designs are emerging. However, their high stiffness-to-mass ratio can often be the cause of noise and vibration problems. Therefore, it is customary to use complex damping materials such as viscoelastic and porous materials in the passive control of structural vibration and noise radiation. The underlying physical mechanisms behind these kinds of damping materials are typically quite complex and highly frequency-dependent, which requires efficient constitutive models able to simply approximate the mechanisms.  Numerical methods are often applied to approximately solve the resulting dynamic equations of complex systems. As the most commonly used predictive tool, the finite element (FE) method is often selected in the design phase to obtain detailed information on the performance of complex vibro-acoustic systems with damping treatments. Its use, however, inevitably leads to very large, complex-valued and frequency-dependent numerical models, which makes original full-order model (FOM) evaluation intractable due to time and memory limitations. Furthermore, the vibro-acoustic optimization problems based on FE model updating require frequently iterative prediction of responses of the large-scale FOM before an optimized solution is arrived, which further increases the computational complexity.In order to ease these problems, model order reduction (MOR) techniques have become indispensable. Most of them, however, do not handle complex-valued and especially frequency-dependent system matrices well. Due to frequency dependencies of the damping material properties, their FE equations of motion are not of a standard second-order form as that for regular elastic FE models. Another practical problem with the use of MOR is the adaptive determination of the dimension of the reduced-order model (ROM) to avoid different trial-errors tests with the FOM for validation. In this dissertation, an adaptive MOR strategy is proposed to reduce the number of degrees of freedom (DOF) involved such that the required computational cost can be largely alleviated while a desired high accuracy can be achieved in a controllable way. This technique consists of making use of Taylor's theorem to the frequency-dependent scalar function(s) coming from the complex material behavior, and then performing the structure-preserving second-order Arnoldi algorithm to solve the underlying FE model in the frequency domain. In support, a relative error indicator is developed to iteratively enrich the reduced model and determine its final order. It should be mentioned that the proposed adaptive MOR process can also be used for standard second-order vibro-acoustic dynamic systems with Rayleigh damping.In addition to rapid prediction of dynamical responses of vibro-acoustic FE systems with damping, the availability of the proposed MOR technique can be exploited in the context of optimization problems with a multi-parameter space. In order to theoretically analyze and design structural systems, the material parameters in the mathematical model of viscoelastic materials need to be known, which can be derived through an inverse identification process. Especially to speed up the exploration of the parametric space, a parametric model order reduction (pMOR) technique is used, where a sampling strategy is introduced. The local orthonormal bases around the selected sampling points are obtained with the proposed adaptive reduction algorithm. A global orthonormal basis can then be constructed by non-weighted singular value decomposition on all local bases. Since the parameter- and frequency dependency can be suitably preserved, the generated single ROM in conjunction with optimization algorithms is very useful to fast identify the material parameters of viscoelastic damping.In order to effectively enhance damping properties under economical- and practical constraints, both the location and the geometry of viscoelastic patches should be optimized. Therefore, the proposed MOR technique is embedded in the explicit moving morphable component (MMC) topology optimization framework to seek the optimal layouts of damping patches under a prescribed area constraint. With the MMC to reduce the number of design variables in the topology formulation and the MOR to reduce the number of DOFs in the FE model, the optimization simulation can be largely sped up." "Adaptive Direct Search for Single- and Multi-Objective Black-Box Optimization" "Wim Desmet" "Mecha(tro)nic System Dynamics (LMSD)" "With the growth of computational power in the last decades, the penetration of numerical methods in engineering workflows has drastically increased. Advanced design and analysis programs have become standard tools in the engineering practice. Concurrently, the state of the art in numerical optimization has also been steadily advancing. Given the close interconnection between engineering and optimization, it is no surprise that the trend towards combining powerful numerical modeling and optimization tools has already started. However, there is still a mismatch between the two disciplines. For one, common assumptions of differentiability and transparency often do not hold for problems based on numerical simulations. Furthermore, it is often impossible to cast a design problem as an optimization with just a single objective.This dissertation aims to address the aforementioned challenges by proposing novel methods for black-box optimization with one or multiple objective functions. The developed methods are fully non-intrusive, require no derivative information, and are robust to irregular phenomena such as hidden constraints and numerical noise. They are designed for efficiency in terms of black-box evaluations and generate feasible iterates at all times, such that restrictions on the computational budget can be easily taken into account. The multi-objective methods proposed in this work approximate the full Pareto front, giving as much information as possible on design trade-offs. On a conceptual level, the algorithms are designed to be understood, implemented, and used with minimal expert intervention.Concerning single-objective optimization, the primary contribution is the gradient-informed generating set search (GIGS) method. GIGS is a directional direct search method for continuous, bound-constrained optimization. The method does not make explicit use of function gradients, but approximates the function topography by forming simplex gradients over the iterations. This information is used to redirect the search, resulting in an adaptive method that offers accelerated convergence without sacrificing robustness. In the present work, the algorithm is first clearly defined and its convergence is analyzed. Then, it is shown that GIGS compares favorably with state-of-the-art direct search methods, both when applied to academic benchmark problems and to realistic engineering use cases.On the front of multi-objective optimization, this dissertation introduces steepest-descent direct multisearch (SD-DMS), an adaptive directional direct search method for continuous optimization with multiple objectives. Like GIGS, SD-DMS exploits the particular form of directional methods to compute simplex gradients without additional function evaluations. Based on the information present in the simplex gradient, the search directions are chosen adaptively so as to attempt improving all objectives simultaneously. In the final part of this dissertation, the SD-DMS algorithms are outlined and several strategies for reducing the overall computational load are provided. In addition, several numerical test cases are reported, in which the performance of SD-DMS is compared to other direct search methods representing the state of the art. " "Sense-IT: Evaluating new sensor technology for energy and growth optimization inornamental crops" "Bert Schamp" "Ornamental plant research" "De doelstelling van het Sense-IT project was via toepassing van sensortechnologie een antwoord bieden op twee vragen die de laatste jaren met regelmaat gesteld worden binnen de sierteeltsector: (1) 'Wat is het effect van mijn teeltmaatregelen en hoe kan ik de groei, ontwikkeling en kwaliteit van mijn gewas nauwlettend opvolgen?' en (2) 'Hoe kan ik het energieverbruik van mijn teeltsturing inschatten en waar mogelijk reduceren?'. Deze vragen zijn een rechtstreeks gevolg van de hoge kwaliteit en kostenefficiëntie die telers nastreven.  Om bovenstaande vragen van de sector te beantwoorden, werden tijdens het kennisopbouwend luik de technologie ontwikkeld en geoptimaliseerd voor de sierteelt (Doel 1). Dit resulteerde in een gevalideerde real-time plantmonitor (standalone applicatie; sensorset + generiek mechanistisch planmodel + PhytoSense Webservice) én een beslissingsondersteunend systeem Sense-IT (real-time plantmonitor + kasklimaatmodel).   Deze ontwikkelingsfase ging gepaard met een grote kennisopbouw rond korte en lange termijn effecten van teeltsturing op de plantrespons en energievraag (Doel 2). O.a. het effect van het verlagen van de stooklijn op plantkwaliteit en energieverbruik (naast reductie van CO2-uitstoot) werden onderzocht.  Kennistransfer rond innovatieve teelt- en klimaatsturing werd gerealiseerd door het vertalen van de opgebouwde kennis naar praktische toepassingen voor telers en toeleveringsbedrijven (Doel 3). Kennis werd naar de sierteeltsector verspreid via publicaties in Sierteelt&Groenvoorziening, Sense-IT brochure, cursussen, workshop en studiedagen. De verworven kennis wordt optimaal verweven in de praktijkwerking van proefcentrum PCS en het verleende individueel advies aan telers. Er werden heel wat adviezen rond teelt- en klimaatsturing gegeven. Op die manier kunnen ook bedrijven die (nog) niet in sensoren investeren toch de vruchten plukken van het Sense-IT project.  Naast semi-praktijk experimenten op diverse sierteeltgewassen, zijn de innovatieve plantsensoren uitgetest op de praktijkbedrijven (Doel 4). De plantmonitor is via een uitleenbare demo-setup op een 25-tal sierteeltbedrijven door de telers zelf uitgetest. Daarnaast werd er ook een offline scenario-calculator gecreëerd waarbij het voorspellend vermogen van Sense-IT gebruikt wordt voor het theoretisch doorrekenen van specifieke vragen van een telers rond teeltstrategie. RESULTATEN & IMPACT: Het gebruik van Sense-IT en de plantmonitor heeft het bewustzijn en de kennis van de telers over de ecofysiologie van zijn gewas zichtbaar vergroot en het gebruik van meer innovatieve energiebesparende teeltstrategieën gestimuleerd. De teler heeft voor het eerst de directe feedback van zijn teeltstrategie op het gedrag van zijn planten kunnen waarnemen, wat een onuitputtelijke bron van nieuwe informatie oplevert. Bovendien zullen de nieuwe teeltstrategieën hem toelaten om het energieverbruik te reduceren zonder kwaliteitsverlies van zijn teelt waardoor zijn competitiviteit verhoogt.  Het economisch voordeel (door reductie energiekost, hogere plantkwaliteit en verkorte teeltcyclus) vormt de motor achter de innovaties en heeft ook een positieve weerklank bij de toeleveringsbedrijven. Tot slot geeft een efficiëntere inzet van energie, reductie van CO2-uitstoot en de verwachtte positieve impuls voor de sector (o.a. behoud van bedrijven) ook een maatschappelijke meerwaarde aan dit project." "ULB-VUB Joint Research Group: Combustion and Robust Optimization - BURN" "Francesco Contino" "Combustion and Robust optimization, Université libre de Bruxelles, Applied Mechanics" "BURN's objective is to contribute to the development of flexible, energy-efficient and non-polluting energy conversion technologies, to fit a continuously changing energy scenario demanding for environmental friendly, secure and cost-effective solutions. BURN promotes the development of a COMBUSTION 2.0 framework,by means of modern strategies based on high-fidelity experimental techniques and numerical simulations, robust optimisation and uncertainty quantification." "COST TU1105 # NVH analysis techniques for design and optimization of hybrid and electric vehicles" "Wim Desmet" "Mecha(tro)nic System Dynamics" "The socio-economic quest towards developing transportation with lower CO2 emission is a global goal of the EU and a crucial ingredient for the competitiveness of the whole European transportation industry. It forces an increased focus on alternative powering systems such as electric and hybrid drives. To be competitive, however, such vehicles must have an acceptable Noise, Vibration and Harshness (NVH) behaviour, not only inside the vehicle, but also outside if it is not to pose major concerns regarding safety of weaker road users such as two-wheelers and pedestrians. Most of the NVH design and problem-solving knowledge gathered has concentrated on internal combustion vehicles and so novel analysis techniques have to be developed for vehicles with these new drives. In addition the limited knowledge on electric and hybrid vehicles is scattered all over Europe. The aim of this COST Action is to engage NVH experts from vehicle industry and renowned research groups in the accumulation, development and dissemination of such novel techniques. The COST framework provides the unique opportunity to bring together experienced academic and early-stage researchers, European authorities for transport regulations, independent consultants, experienced representatives from industry and associations of transporters." "Optimization of the functioning of the housing inspectorate in relation to the housing actors." "Wouter Van Dooren" "Public Administration & Management" "The research project studies the role of the housing inspectorate in relation to the housing actors (housing corporations, social rental agencies, credit agencies, ...). The purpose is to optimize the effectiveness of the inspectorate within the current regulatory framework." "Optimization of RFID Systems in Unfriendly Operating Environments" "Bart Nauwelaers" "ESAT- TELEMIC, Telecommunications and Microwaves, Electrical Engineering Technology (ESAT), Ghent and Aalst Technology Campuses" "This work focused on HF RFID systems, where it was a goal to increase the reliability of the system. Especially in environments with metallic parts many problems occur. It became clear that many problems took place at the loop antenna of the RFID system. Therefore, we focussed on the design of loop antennas in order to increase the reliability of HF RFID systems. Although HF RFID systems are widely adopted by the industry, it was shown that the implementation of an RFID system for a specific environment involves a lot of design choices. Considering the loop antenna, several important design parameters have to be taken into account if we are aiming for a proper functioning HF RFID system. Metal objects in the vicinity of the loop antenna can have a high impact on the functionality of the system. In some cases a redesign of the loop antenna is needed. A redefinition of several design parameters and a solution, in terms of a new antenna concept, was proposed to take care of metal objects in the vicinity of the antenna. It has been made clear that an environment with metal objects has to be handled with care. If we only consider simple environments with one metal plate or object, we can design the loop antenna analytically. But when the environment becomes more complex, one needs a strong tool to create a reliable loop antenna for these environments. The main result of this thesis is the creation of an automated antenna design tool (AADT) which takes care of metals in the vicinity and creates an optimized loop antenna. The optimized loop antennas are based on parametric shapes and self-defined RFID goal functions. Two practical and challenging test cases served as an evaluation of the AADT in which an optimized loop antenna was implemented. In both cases, we created a reliable HF RFID system, which always met the specified requirements. At all time, measurements confirmed the simulation results." "Data-driven distributed control and optimisation for multi-energy demand management in local energy communities and microgrids" "Geert Deconinck" "Electrical Energy Systems and Applications (ELECTA)" "Multi-energy systems use multiple energy carriers (electricity, heat, gas,…) to supply end-users with energy services such as space heating, hot water, lighting, electric vehicle charging, etc. These energy carriers are coupled to each other via energy conversion units (coupled heat power generation, heat pumps, electric and gas boilers, etc.). The ability to supply the energy service from different carriers allows for operational flexibility, which can be used to optimise objective functions such as minimising primary energy use, minimising cost, maximising profits for different actors, etc. The operational freedom to exploit the flexibility only becomes larger if storage devices (hot water storage tanks, batteries, seasonal thermal storage, etc.) are added. Such storage devices are often shared between different households in a residential context, or between different end users in a more industrial context, leading to local energy communities and microgrids. At this level, also renewable energy sources (photovoltaics or wind turbines) are added to reduce grid dependence. Thanks to the addition of information and communication technology much sensor data is available about the status of the grids and its connected devices (three-phase voltages and currents for the electrical grids, temperatures and flows for the heat grids, state-of-charge of batteries and storage tanks, energy requirements of controllable devices, such as electric vehicle charging requirements or hot water use profiles, etc.), and many devices in the system can be controlled – resulting in smart, multi-energy systems. Due to the uncertainties related to user behaviour, weather change or energy prices, it is challenging to do planning and operations of the controllable devices in these local energy communities and microgrids. Thanks to the large amounts of sensor data (both in real-time and historic), it is possible to forecast the uncertainties (using machine learning techniques) and to use data-driven control methods for the controllable devices, and for system optimisation. A combination of the local flexibility control and distributed control methods is needed to deal with the hierarchical requirements in a microgrid, needing a balance between demand and supply over all energy carriers. Such distributed control can be completely decentralised with multi-agent systems, or be hierarchically structured in a more classical approach. Finally, this control must incorporate multiple time dimensions (seconds, minutes, hours, days, months), due to the different time constants involved in electrical and thermal balancing of generation, load and storage devices, leading to multi-scale, multi-energy systems. This leads to the central research question in this PhD thesis, i.e. how to use data-driven, distributed methods to control the controllable storage, generation and end-use devices, such that optimal energy services are delivered in a microgrid context."