"uuid","repository link","title","author","contributor","publication year","abstract","subject topic","language","publication type","publisher","isbn","issn","patent","patent status","bibliographic note","access restriction","embargo date","faculty","department","research group","programme","project","coordinates"
"uuid:8cf78935-a0d6-42d6-a1fc-9e25ce285fa9","http://resolver.tudelft.nl/uuid:8cf78935-a0d6-42d6-a1fc-9e25ce285fa9","Supporting non-expert users in modelling and understanding AI, an interactive CP approach: Bringing the power of advanced optimisation in employee scheduling to small and medium-sized organisations","Kropf, Kylian (TU Delft Electrical Engineering, Mathematics and Computer Science)","Yorke-Smith, N. (mentor); Bormans, R. (graduation committee); Tielman, M.L. (graduation committee); Delft University of Technology (degree granting institution)","2024","This thesis proposes and develops an interface and model in which advanced optimisation for general employee scheduling is made available to non-experts in computer science or optimisation. The interface teaches, guides, configures, dynamically creates a constraint programming (CP) model, iteratively improves, decreases black box properties, increases trust in the outcome, and complies with relevant European Union Artificial Intelligence law. The objective of this study is to allow a wider range of organisations to take advantage of CP techniques, with the potential to greatly improve efficiency, reduce unfairness, meet company goals, and improve employee satisfaction.
Employees are assigned to personalised shifts based on the expected demands of departments within an organisation, these are set by the domain expert in the field of employee scheduling. Next, to tailor the model to the organisation’s needs, the domain expert is guided in setting both the restrictive assumptions and priorities of shift assignment. To optimise the generated optimal schedule, it is encouraged to create and compare multiple sets of configurations and subsequent schedules. Historical demand data, organisation structure, contract information, and employee preferences are included. Multiple visual design iterations have been made, after which a working interface has been developed and improved iteratively in conjunction with experts in the field. Validations with external domain experts from various industries and organisations have shown that the interface performed effectively in supporting the objectives. Further research can be done to improve the speed of solving, implement diversity of solutions, support for highly custom constraints based on natural language, or interface reusability for other optimisation problems.","combinatorial optimisation; multi objective optimisation; general personnel scheduling problem; domain expert systems; understandable AI; XAI","en","master thesis","","","","","","https://dyflexis.com Unspecified Dyflexis Employee Scheduling Software","","2024-12-31","","","","Computer Science","",""
"uuid:c469a4fa-92bc-4ddf-9790-bd9f7a248815","http://resolver.tudelft.nl/uuid:c469a4fa-92bc-4ddf-9790-bd9f7a248815","Dynamic Wireless Charging of Electric Vehicles","Shi, W. (TU Delft DC systems, Energy conversion & Storage)","Bauer, P. (promotor); Dong, J. (copromotor); Delft University of Technology (degree granting institution)","2023","As a more convenient alternative to conductive charging technology, wireless charging is seen as a key technology drive for transportation electrification. In electric vehicle (EV) battery charging applications, wireless power is transferred through a magnetic link, so it is referred to as inductive power transfer (IPT). One advantage of IPT technology is that the charging of EVs can be fully automated. The recharge of traction batteries can start automatically when the EV stops where its receiver (Rx) coil is coupled with a transmitter (Tx) coil of an IPT charger. Apart from static charging applications, IPT technology can also be used to build dynamic charging roads where EVs can get charged in motion and the capacity of onboard batteries can be reduced. This thesis studied four challenges that should be addressed before dynamic IPT becomes mature enough for commercial use. The research topics focus on magnetic coupler design, prediction and control of transient behaviors, reduction of power fluctuation, and detection of EVs and foreign objects (FOs).
Magnetic coupler design
The key performance indicators of an IPT system include power transfer capability, power density, power efficiency, and misalignment tolerance. Due to conflicts among these performance indicators, it is indispensable to formulate the design of IPT charging pads as a multi-objective optimization (MOO) problem. By using finite element (FE) models, the magnetic field property of a coupler can be computed. However, calculating the aligned and misaligned power losses at the rated power requires not only the magnetic field property but also the compensation strategy. The compensation strategy determines the load match method which is used to calculate the optimal load condition and the rated winding currents. Therefore, compensation strategy should also be considered for the magnetic coupler design. With the magnetic field distribution known, the power losses in the AC link can be calculated through the existing analytical method.
This thesis develops a MOO method that can find the performance space from the design search space of magnetic couplers. In the performance space, Pareto fronts can be obtained under different conflicting optimization objectives. The study shows that analytically calculating the AC link power efficiency is possible when the magnetic field is accurately computed at the rated condition. More importantly, the DC-DC power efficiency of the final prototype reaches $97.2\%$ which proves that the MOO design is vital to make full use of IPT technology.
Prediction and control of transient behaviors
IPT systems require capacitive/inductive components to form resonant circuits on both sides to improve the power transfer capability and power efficiency, while the compensation components also make the resonant stage of a high order. As a result, the analytical dynamic models of IPT systems are complex and mostly impossible to solve in the time domain.
This thesis proposes a new reduced-order dynamic modeling method that describes the transient behavior of a resonant stage from the energy point of view. The order of the resultant dynamic model is one-fourth that of conventional ones for SS compensated IPT systems. Also, a MPC controller is designed based on the proposed dynamic model. It is proven that simplifying the dynamic model is helpful in explaining how circuit parameters influence transient behaviors and also in facilitating the application of advanced control strategies in IPT systems.
Reduction of power fluctuation
The most obvious difference between static and dynamic IPT is the change in magnetic coupling. In DIPT applications, the magnetic coupling fluctuates from the maximum to a usable level as EVs move, so one of the main challenges of DIPT is to stabilize the pick-up power, especially for DIPT systems using segmented Tx coils where magnetic coupling changes more frequently. The conventional methods are either to overlap Tx coils or to add extra sets of the Rx sides, which are expensive in building costs.
This thesis presents the design of a segmented DIPT system using a multiphase Tx side. The Rx coil consists of two sub-windings connected in series with a relatively large spatial offset in the EV moving direction. One advantage of the proposed design is that the Tx coils are deployed loosely so the building cost can be reduced. The other advantage is that the pick-up power is seamless with a small ripple. The pick-up power demonstrates a $24.9\%$ ripple by experiments.
Detection of EVs and FOs
To minimize the Tx side power losses and magnetic field radiation, the detection of EVs and FOs should be implemented in DIPT systems. Considering the integration of the detection equipment into the charging pads, PCB coils become the most suitable candidate to sense the magnetic field for detection purposes. However, the detection of EVs and FOs are mostly discussed separately in the literature. There is a need to achieve these two detection functions within one set of PCB coils.
This thesis presents the design of detection equipment consisting of PCB coils installed onto charging pads and the detection resonant circuit (DRC) connected to Tx side PCB coils. It can be concluded that the detection of EVs and FOs can both be realized by measuring the variation of the magnetic field caused by their intrusion, and PCB coils demonstrate good performances in measuring the change of magnetic field together with DRC to amplify the detection signals.
In order to identify the impacts of the Hyperloop network design in the global transportation sector, a literature review was conducted on the transformative potential of the Hyperloop. Key strengths were identified as a reduction in travel times and low operational emissions. On the other hand, the high capital resources required and the uncertainty around the safety of technology were the main points of criticism. In order to analyze the potential demand for Hyperloop and model the modal shift, a Multi- Nominal Logit was employed where a utility function was formulated for the total benefit passengers receive upon completing a trip. The key attributes for the utility function were selected as travel time, travel costs, number of transfers, and safety perception, in alignment with previous studies on the subjects. A utility-based probabilistic mode choice was determined for the available demand. A multi-objective optimization problem was formulated for the facility-location network design of Hyperloop.
The decision variables of the model were formulated as the decision to open a Hyperloop hub at a location and the decision to build infrastructure between the selected Hyperloop hubs. The model output is an alternate network optimized for four different objective functions. These objectives are determined to be (1) Utility Maximization, (2) Probability of Purchase Maximization, (3) Emission Minimization, and (4) Revenue Maximization as these factors were determined to be key performance indicators in a prospective Hyperloop network. The model aims to provide the decision-makers with an overview of the trade-offs involved with varying objective criteria considered in the network generation.
A case study was created to test the model within Europe. The main aim of the case study is to assess the economic and environmental impacts of the Hyperloop system and provide recommendations to policymakers regarding the conception of the Hyperloop network within the European Union. The case study employs the NUTS classification and excludes European countries where the demand data is incomplete and focuses on countries within the TEN-T network. Furthermore, three categories of experimental scenarios were set up to assess the sensitivity of the model to parameter values. The categories are (1) pricing strategy scenarios, (2) safety perception scenarios and (3) policy intervention scenarios. The findings reveal significant disparities in network characteristics based on different objective criteria. The Utility Maximization objective focuses on maximizing trip utility, leading to a network design with direct links between hubs, resulting in compact networks and lower infrastructure costs. However, Spain and Italy have lower priority in this design. On the other hand, the other three objectives (probability of purchase maximization, emission minimization, and revenue maximization) yield networks with a minimum-spanning tree pattern. These networks outperform the utility maximization network in terms of attracting passengers, reducing emissions, and economic performance. To maximize societal benefits, it is recommended to prioritize the remaining three objectives. The study finds that Hyperloop becomes more competitive for longer-distance trips. Experimentation with ticket prices, safety perception, and policy interventions demonstrates their influence on modal share, revenue stability, and carbon emissions. Higher ticket prices discourage Hyperloop usage, safety perception plays a crucial role, and policies discouraging short-haul flights result in higher Hyperloop modal share and lower emissions. These findings highlight the importance of considering ticket prices, safety perception, and strategic policies to promote sustainable transportation and reduce carbon emissions through a modal shift to Hyperloop.
Future research opportunities include expanding the utility function to incorporate additional attributes affecting mode choices, exploring modal shifts from other modes to Hyperloop, relaxing assumptions about geographical obstacles and hub locations, integrating strategic and tactical planning, and validating the model with a broader range of origin-destination pairs. Computational performance can be enhanced using meta-heuristics to compare different heuristics for network outputs and efficiency.
Previous research has highlighted how an off-grid configuration would result in inconveniently high costs for the community's users, if compared to the average cost of energy in The Netherlands. The aim of this thesis is to study the system in a grid-connected configuration, and in particular to find the optimal sizes of the components in order to achieve the best trade off between three conflicting objectives : minimizing total costs, maximizing self- sufficiency and maximizing reliability. After modeling the system's components and their mutual interactions, the optimization was carried out on MATLAB using a variant of the NSGA-II algorithm, which provides a Pareto Set of equally optimal solutions for the problem. The solutions were then ranked with a Technique for Order Preference based on Similarity to the Ideal Solution (TOPSIS), to assist the decision-making process.
The simulations have determined that an installed capacity of 85.41 kWp (composed of 234 panels of 365 Wp each) results in the most effective choice for the solar energy generation, irrespective of the external conditions imposed. The optimal storage capacity, however, results significantly more influenced by factors such as grid imports limitations and price uncertainties. Under the conditions of limited imports from the grid, an optimal capacity of 75 kWh in the form of batteries was found. In general, the study confirms that the adoption of an hydrogen storage system is far from being convenient on a small scale residential level, regardless of the pricing conditions. The research has also posed an accent on the incremented costs incurred to reach full reliability of the system with low values of dependence from the grid, due to the high costs of the necessary storage equipment. Additionally, despite the best solutions found represent the optimal compromises balancing the conflicting objectives, reasonable solutions in terms of costs faced by the Community's users are usually not among the first choices of the ranking algorithm, mainly because they necessitate of at least 50% of the load to be supplied through grid imports.","Energy Community; Multi Objective Optimisation; PV system; Hydrogen storage","en","master thesis","","","","","","","","2025-09-06","","","","Electrical Engineering | Sustainable Energy Technology","24/7 Energy Lab",""
"uuid:94e6fc04-ff15-4fa1-96cd-99248f1eaf61","http://resolver.tudelft.nl/uuid:94e6fc04-ff15-4fa1-96cd-99248f1eaf61","A surrogate-assisted propeller optimisation in propulsive and regenerative operations: A case study","van Heugten, Lars (TU Delft Mechanical, Maritime and Materials Engineering)","van Terwisga, T.J.C. (mentor); Foeth, E.J. (mentor); Scholcz, T.P. (mentor); Delft University of Technology (degree granting institution)","2023","Engineering has become more and more optimisation and simulation based. This causes an increase in the use of optimisation methods for complex problems. This thesis focuses on the implementation of a surrogate-assisted optimisation method for propeller optimisation, to design a better propeller in a shorter timespan. Therewith reducing both: computational expenses, and fuel consumption when a propeller has been built. This thesis tries to reach two objectives: The first goal is to implement a surrogate-assisted method for propeller optimisation. The other objective is to design a propeller for both propulsive as well as for regenerative operations. To succeed, this thesis has been limited to the testing and coupling of SAMO-COBRA (A Fast Surrogate Assisted Constrained Multi-objective Optimization Algorithm) with PropArt. The testing has been based on benchmark tests where SAMO-COBRA was tested against CMOPSO and NSGA-II (the two algorithms that MARIN currently uses for propeller optimisation). Here, SAMO-COBRA outperformed NSGA-II and CMOPSO on over half of the test cases and is therefore considered for the rest of the research.
The thesis evaluates propellers using a Boundary Element Momentum theory, which is a mathematical method that is the basis of MARINs in-house propeller tool PROCAL. To be able to model propellers that operate at high J values correctly certain improvements regarding the wake expansion and alignment have been made. For the wake expansion three methods are proposed. After comparing the open water diagram of a F4-63-0.6 propeller, that is evaluated by PROCAL, and the one that was made after physical testing. It can be concluded that the disk theory is the most versatile implementation for propellers that operate at a range of advance ratio. Results show that from the two proposed methods, the method where the wake pitch is prescribed by the advance ratio results in the best wake alignment.
The test case optimises 3 configurations, two separate designs, one for propulsion and one for regeneration. A normal CPP that is loaded under a negative angle of attack for regeneration and a propeller that can be fully reversed for regenerative operations (the propeller is loaded from the trailing edge in regeneration). The propeller is optimised to reach a maximum regenerative power whilst minimising the propulsive power. The optimum propeller has to satisfy multiple constraints based on cavitational-, geometrical- and separation of flow constraints. The optimisation using SAMO-COBRA did not yield feasible results for any of the three cases. After removing the cavitational constraints a new analysis of the results has been done, it is found that a propeller that is rotated 180 degrees can provide the highest regenerative power. This thesis proposes multiple hypotheses that cause the lack of feasible results of the test case. The proposed causes are: errors in the PropArt model, problems in the compiled PropArt version or a too low convergence level of COBYLA.
Today Boerhaavewijk is showing its age. The modernist principles have become old-fashioned and do not support a livable neighbourhood. The buildings do not comply with today's standards and are rapidly becoming outdated in both their function and technical state. The local architecture and urban planning don’t foster social cohesion or a strong identity.
Over time the demographics have been changing and growing, with a more diverse population as a result. Different migrant communities have made Boerhaavewijk their home. Spaces where people can meet however are limited in quantity and quality. There is a lack of future-proof public facilities and community spaces.
The current demand for housing has resulted in a lot of new construction in the neighbourhood, often replacing the old. The same construction methods and materials such as concrete, steel and brick are used, which have a large carbon footprint and are high in embodied energy. These buildings are not designed to be flexible and re-used, guaranteeing their demolition in the future. We use the same non-renewable materials and unsustainable construction methods as in the 60s.
In order to tackle these problems, the graduation project proposes the construction of Forum Boerhaave: the design of an iconic and sustainable building in Boerhaavewijk that can adapt to demographic changes. The forum is a multifunctional public place for people of the neighbourhood to meet, a community building to facilitate its users.
The site of the building is found in between the urban and the natural, connecting Boerhaavewijk and the Poelpolder. The building attempts to find a balance between the generic and the specific, as its intention is to facilitate instead of dictate, while also reacting to the specific conditions or the genius loci of the site.
In order to future-proof the building, the ‘open building’ principles of Habraken are applied. This way the building may adapt to future changes when demanded by its users. The materials used in the building are, as far as possible, locally harvested. An open plan is achieved through a timber skeleton structure, CLT floors and an aluminium space frame, the latter reclaimed from Schiphol. The plinth is 3D printed, using dredged earth from the river Spaarne. It is stabilised using Kaumera biopolymer and insulated with cellulose both harvested from sewage treatment plants, while reed fibres from the Poelpolder improve the tensile strength of the material.","Open Building; Harvest; Post 65; Biobased; Circular Design; Additive Manufacturing; Earth Construction; Multi Objective Optimisation","en","master thesis","","","","","","","","","","","","Architecture, Urbanism and Building Sciences | Architectural Engineering","","52.364464, 4.665682"
"uuid:90723fc9-e862-448d-aa7c-9ef8795c35a7","http://resolver.tudelft.nl/uuid:90723fc9-e862-448d-aa7c-9ef8795c35a7","Parallel cost-aware optimization of multidimensional black-box functions","Sihlovec, Oliver (TU Delft Electrical Engineering, Mathematics and Computer Science; TU Delft Algorithmics)","Spaan, M.T.J. (mentor); de Vries, J.A. (mentor); Lofi, C. (graduation committee); Delft University of Technology (degree granting institution)","2023","Scientific problems are often concerned with optimization of control variables of complex systems, for instance hyperparameters of machine learning models. A popular solution for such intractable environments is Bayesian optimization. However, many implementations disregard dynamic evaluation costs associated with the optimization procedure. Furthermore, another common trope among
Bayesian algorithms is that they are short-sighted and do not consider long-term effects of their actions. This paper investigates the viability of multitimestep cost-aware Bayesian optimizers and evaluates their performance in environments with delayed rewards. To this end, we combine existing works on parallel Bayesian optimizers and costaware heuristics. Our findings reveal that although
such parallel optimizers yield more optimal results and are more resistant to delayed feedback compared to their myopic counterparts, they are unable to achieve cost-awareness.","Bayesian Optimization; Machine learning; Multi Objective Optimisation; Parallelization; Cost-awareness","en","bachelor thesis","","","","","","https://github.com/osihlovec/ca-qei This link redirects to the student's code repository containing the code, which has been utilized throughout the research and which can be used to reproduce the results outlined in the paper.","","","","","","Computer Science and Engineering","CSE3000 Research Project",""
"uuid:0e03913c-898e-4392-8de5-072a7ead7fd6","http://resolver.tudelft.nl/uuid:0e03913c-898e-4392-8de5-072a7ead7fd6","Optimal Mixing Evolutionary Algorithms for Large-Scale Real-Valued Optimization: Including Real-World Medical Applications","Bouter, P.A. (TU Delft Algorithmics; Centrum Wiskunde & Informatica (CWI))","Bosman, P.A.N. (promotor); Alderliesten, T. (copromotor); Delft University of Technology (degree granting institution)","2023","In recent years, the use of Artificial Intelligence (AI) has become prevalent in a large number of societally relevant, real-world problems, e.g., in the domains of engineering and health care. The field of Evolutionary Computation (EC) can be considered to be a sub-field of AI, concerning optimization using Evolutionary Algorithms (EAs), which are population-based (meta-)heuristics that employ the Darwinian principles of evolution, i.e., variation and selection. Such EAs are historically mainly considered for the optimization of difficult, non-linear problems in a Black-Box Optimization (BBO) setting, because EAs can effectively optimize such problems even when very little is known about the optimization problem and its structure. This is in contrast to optimization methods that are specifically designed for certain problems of which the definition and structure are known, i.e., a White-Box Optimization (WBO) setting.","Evolutionary Algorithms; Gene-pool Optimal Mixing; Gray-box optimization; Large-scale optimization; Real-valued optimization; Multi-objective Optimisation; Graphics Processing Unit (GPU); CUDA; Brachytherapy; Treatment planning; Deformable image registration","en","doctoral thesis","","978-94-6366-648-0","","","","","","","","","Algorithmics","","",""
"uuid:68e873db-82af-49e0-8b53-85de2570f09b","http://resolver.tudelft.nl/uuid:68e873db-82af-49e0-8b53-85de2570f09b","On the Strengths of Pure Evolutionary Algorithms in Generating Adversarial Examples","Bartlett, A.J. (TU Delft Multimedia Computing); Liem, C.C.S. (TU Delft Multimedia Computing); Panichella, A. (TU Delft Software Engineering)","","2023","Deep learning (DL) models are known to be highly accurate, yet vulnerable to adversarial examples. While earlier research focused on generating adversarial examples using whitebox strategies, later research focused on black-box strategies, as models often are not accessible to external attackers. Prior studies showed that black-box approaches based on approximate gradient descent algorithms combined with meta-heuristic search (i.e., the BMI-FGSM algorithm) outperform previously proposed white- and black-box strategies. In this paper, we propose a novel black-box approach purely based on differential evolution (DE), i.e., without using any gradient approximation method. In particular, we propose two variants of a customized DE with customized variation operators: (1) a single-objective (Pixel-SOO) variant generating attacks that fool DL models, and (2) a multi-objective variant (Pixel-MOO) that also minimizes the number of changes in generated attacks. Our preliminary study on five canonical image classification models shows that Pixel-SOO and Pixel-MOO are more effective than the state-of-the-art BMI-FGSM in generating adversarial attacks. Furthermore, Pixel-SOO is faster than Pixel-MOO, while the latter produces subtler attacks than its single-objective variant.","Black-box testing; Adversarial example generation; Differential evolution; Multi-Objective Optimisation; Search-based Software Testing; Deep Learning","en","conference paper","IEEE","","","","","Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.","","2024-01-27","","","Multimedia Computing","","",""
"uuid:66372140-4a90-4538-ae15-79fe444033c2","http://resolver.tudelft.nl/uuid:66372140-4a90-4538-ae15-79fe444033c2","Multi-objective Optimisation Framework for Assessment of Trade-Offs between Benefits and Co-benefits of Nature-based Solutions","Yang, Shengnan (IHE Delft Institute for Water Education); Ruangpan, L. (TU Delft BT/Environmental Biotechnology; IHE Delft Institute for Water Education); Torres, Arlex Sanchez (IHE Delft Institute for Water Education); Vojinovic, Zoran (University of Belgrade; University of Exeter; IHE Delft Institute for Water Education)","","2023","Urbanization and climate change are producing an escalation in the prevalence of urban problems, particularly those connected to flooding, prompting authorities and stakeholders to recognize the need for sustainable solutions. Nature-Based Solutions are progressively replacing traditional engineering solutions as an alternative since they are more eco-friendly. By re-activating the urban hydrological cycle processes, NBS intends to increase the natural water storage capacity to help decrease urban flooding. The work described here outlines a framework for optimising the efficacy of NBS for flood risk reduction and its co-benefits, as well as defining the trade-offs among these co-benefits. The framework integrates 1D hydrodynamic models with multi-objective optimisation techniques. To demonstrate the applicability of the framework and its methods it has been used in Sint Maarten, which is an island located in the Caribbean Sea. Four NBS measure were identified as having good potential to be applied in the case study, namely: green roof, permeable pavement, bio-retention pond, and open detention basin. The results showed that the developed framework has the ability to represent the link between benefits and costs when evaluating various NBS, hence aiding the decision-making process to select and implement NBS.","Flood risk reduction; Multi-objective optimisation; Nature-based solution; NBS benefits; NSGA-II; SWMM model","en","journal article","","","","","","","","","","","BT/Environmental Biotechnology","","",""
"uuid:1d37e009-7593-4848-a4ca-c6e58a575704","http://resolver.tudelft.nl/uuid:1d37e009-7593-4848-a4ca-c6e58a575704","Parameter space exploration for the probabilistic damage stability method for dry cargo ships","Milatz, Bas (Student TU Delft; C-Job Naval Architects); de Winter, Roy (Universiteit Leiden; C-Job Naval Architects); van de Ridder, Jelle D.J. (C-Job Naval Architects); van Engeland, Martijn (DELFTship Maritime Software); Mauro, F. (TU Delft Ship Design, Production and Operations); Kana, A.A. (TU Delft Ship Design, Production and Operations)","","2023","The prediction of the statutory attained subdivision index is a challenging issue for the initial design of ships due to the design freedom offered by a probabilistic damage stability assessment. To this end, optimisation techniques integrated with a parametric model of the internal layout may generate a preliminary subdivision design, fulfilling damage stability regulations and cargo volume requirements. The present study explores using a multi-objective constrained optimisation algorithm coupled with a parametric model of a single hold cargo vessel, first investigating two design goal alternatives and secondly performing a global sensitivity analysis on the design variables for the most promising solution. The adoption, in parallel, of state-of-the-art practices shows the validity of the obtained solutions and the time benefits for designers. Nonetheless, the non-linear nature of probabilistic damage stability does not allow for clearly identifying the most impactful parameters on the attained survivability index.","Damage stability; Global sensitivity analysis; Multi-objective optimisation; Parametric model; Ship design","en","journal article","","","","","","","","","","","Ship Design, Production and Operations","","",""
"uuid:3fdfdf01-45f9-4cb9-ac54-ad98edae408c","http://resolver.tudelft.nl/uuid:3fdfdf01-45f9-4cb9-ac54-ad98edae408c","Optimizing machine learning inference queries for multiple objectives","Schönfeld, Mariette (TU Delft Electrical Engineering, Mathematics and Computer Science)","Houben, G.J.P.M. (mentor); Katsifodimos, A (graduation committee); Hai, R. (graduation committee); Delft University of Technology (degree granting institution)","2022","Machine learning inference queries are a type of database query for databases where a model pipeline is needed to evaluate its boolean predicates. Using a model zoo it is possible to select a variety of models to execute in a sequence rather than using a highly specialized model to answer every query predicate. Machine learning models can have multiple measurements for gauging performance however, and the quality of a query plan therefore is not only dependent on the time needed to compute it. Selecting a query plan of models that balances multiple objectives is not a trivial feat however. This work builds upon existing methods that utilize MIPs for model selection and ordering for machine learning inference queries by extending them with multi-objective optimizing capabilities. The opportunity for adding a third objective, namely memory footprint, to that of accuracy and execution cost is explored. Several methods are then considered and compared on their suitability, and the final chosen method, the Archimedean goal method, can generate Pareto optimal query plans that provide gains over naive, greedy methods. In addition, several methods of cutting down runtime on the original optimizer are explored, leading to a program than can generate higher quality solutions in less time.","Query optimization; Machine learning; Multi Objective Optimisation","en","master thesis","","","","","","","","","","","","","",""
"uuid:389e6a70-ffb6-45cd-b97a-ed259daba324","http://resolver.tudelft.nl/uuid:389e6a70-ffb6-45cd-b97a-ed259daba324","Sizing Optimization of a Hybrid Propulsion powertrain for a Crew Transfer Vessel incorporating uncertainties: towards a cost-effective, eco-friendly design","Karagiorgis, Savvas (TU Delft Mechanical, Maritime and Materials Engineering)","Polinder, H. (mentor); Wang, X. (mentor); Nasiri, S. (mentor); van Biert, L. (graduation committee); Delft University of Technology (degree granting institution)","2022","A major transition became necessary for the maritime industry to meet the IMO’s targets for mitigating the carbon footprint of the sector by at least 50% by 2050. One of the promising methods to lower the emissions is the ship's hybridization. An enormous increase in pilot and demonstration projects for that purpose is being observed. This research was contacted through the involvement of TU Delft in such a project called the Implementation of Ship Hybridisation.
The design and optimization of hybrid propulsion systems is a complex and challenging task due to the different power sources involved and the dependence on the energy management and control. The physical system and the control algorithm should be designed in an integrated manner to obtain an optimal system design. This study applies a multi-objective double-layer optimization methodology to optimize the sizing and energy management of a hybrid ship propulsion system to be installed on a Crew transfer vessel. A proposed hybrid topology which combines diesel engines, batteries and fuel cells is considered. The proposed approach incorporates the development of fuels and electricity prices as well as the investment costs of the system’s components as an uncertainty element. The introduction of emission reduction measures such as carbon tax was also considered in the study. Future trajectories for the relevant uncertainties were developed and incorporated in the optimization methodology to provide decision-makers with a more realistic picture of the solution space.
The analysis of the optimization results was based on the Total cost of ownership (TCO) and the emission reduction potential of the optimal designs produced by the optimization methodology. The results show that instead of choosing a hybrid propulsion system for the vessel under study, an all-electric propulsion system which is based entirely on batteries and fuel cells is the most economical and environmentally friendly option. Incorporating diesel engines has a negative impact on the operational expenditures of the system in the long term. A fully electric propulsion system would require a larger initial investment from the ship owner, but it would pay off over the course of the ship’s remaining useful life.
This research can be used as a reference to base the decision of the stakeholders on choosing a new propulsive system for their vessel. The parameters of the optimization methodology can be easily changed to explore more options and expand the design solution space if this thesis’ results don’t satisfy the shipowner. In addition, it is suggested that a professional user interface designer be involved in the development of a real life decision support tool, incorporating the multi-objective optimization methodology.
There are numerous commercial software available that provide these functionalities. However the methods used to model machines within these packages are not available or cannot be changed. They are also usually expensive. This makes the exploration of new limits and new topologies difficult. Using open source packages allows us to modify these methods according to our application and gives greater control over the process.
This thesis uses PYLEECAN python library that is based on FEMM software to perform the analysis of the machine. A six time-step magnetostatic analysis method to calculate the average torque and iron losses in the stator and rotor core is presented along with methods to calculate the copper losses and windage losses. A steady state Lumped Parameter Thermal Network (LPTN) is developed to calculate the temperatures of various parts of the machine. The LPTN is capable of estimating the temperatures under natural convection and forced air cooling conditions.
Finally a MOO framework was developed using the models developed and the PYMOO python library. This thesis uses NSGA-II to perform the MOO. The MOO framework was used to optimise the design of a machine for a drone application and explore the specific power density limit of the machine. The power density limit was found to 5 − 7 kW/kg based on different slot pole combinations, winding temperature limits and core material used. Further, insights into how different machine parameters affect the specific power density are presented
Next the development of the model and solution method, the implemented method is used to investigate the collaborative gain for four levels of collaboration in a dynamic environment. The level of collaboration is expressed by the number of requests that is reassigned between carriers. The results show that up to 23% collaborative gain can be realized in the dynamic setting.","Collaborative gain assesment; Dynamic Vehicle Routing Problem; Horizontal collaboration; Multi Objective Optimisation","en","master thesis","","","","","","","","","","","","Marine Technology | Transport Engineering and Logistics","",""
"uuid:44a827dc-85b3-43a2-84c4-274e016913bb","http://resolver.tudelft.nl/uuid:44a827dc-85b3-43a2-84c4-274e016913bb","Exploring trade-offs among the multiple benefits of green-blue-grey infrastructure for urban flood mitigation","Alves, Alida (TU Delft BT/Environmental Biotechnology; IHE Delft Institute for Water Education); Vojinovic, Zoran (IHE Delft Institute for Water Education); Kapelan, Z. (TU Delft Sanitary Engineering; University of Exeter); Sanchez, Arlex (IHE Delft Institute for Water Education); Gersonius, Berry (ResilienServices)","","2020","Climate change is presenting one of the main challenges to our planet. In parallel, all regions of the world are projected to urbanise further. Consequently, sustainable development challenges will be increasingly concentrated in cities. A resulting impact is the increment of expected urban flood risk in many areas around the globe. Adaptation to climate change is an opportunity to improve urban conditions through the implementation of green-blue infrastructures, which provide multiple benefits besides flood mitigation. However, this is not an easy task since urban drainage systems are complex structures. This work focuses on a method to analyse the trade-offs when different benefits are pursued in stormwater infrastructure planning. A hydrodynamic model was coupled with an evolutionary optimisation algorithm to evaluate different green-blue-grey measures combinations. This evaluation includes flood mitigation as well as the enhancement of co-benefits. We confirmed optimisation as a helpful decision-making tool to visualise trade-offs among flood management strategies. Our results show that considering co-benefits enhancement as an objective boosts the selection of green-blue infrastructure. However, flood mitigation effectiveness can be diminished when extra benefits are pursued. Finally, we proved that combining green-blue-grey measures is particularly important in urban spaces when several benefits are considered simultaneously.","Flood damage reduction; Hybrid drainage infrastructure; Multi-objective optimisation; Multiple benefits; Nature based solutions","en","journal article","","","","","","","","","","","BT/Environmental Biotechnology","","",""
"uuid:516aa463-3d7a-4d65-8092-6d0ef7a61562","http://resolver.tudelft.nl/uuid:516aa463-3d7a-4d65-8092-6d0ef7a61562","Selection of cost-effective emission abatement options for early-stage ship design: A selection tool implemented for a road ferry and workboat","van Grootheest, Ivar (TU Delft Mechanical, Maritime and Materials Engineering)","Pruijn, Jeroen (mentor); Frouws, Koos (graduation committee); Duinkerken, Mark (graduation committee); Delft University of Technology (degree granting institution)","2019","The maritime emission regulations and incentives require and motivate the selection of different types of energy systems, fuels, and abatement options. In early-stage ship design, the design space of possible combinations can be significant. Therefore, a selection tool has been developed to find the most satisfactory combination of abatement options at minimum costs that at least meet the emission requirements. This decision problem is studied from a general perspective to develop a universal selection tool to enable a widespread application in the maritime industry. The selection tool contains datasets with decision parameters of different energy systems, fuels, and abatement options. The energy systems, including reference fuels, can be selected and assessed on their annual economic and environmental performance. Both upstream and operational emissions are considered and, in addition, the external costs of the emissions are quantified. The abatement options, including fuels, have different effects on fuel consumption and emissions. Moreover, interaction effects occur when combining alternatives. The identified problem is a combinatorial optimisation problem and the objective space is constrained by the emission and compatibility constraints. It is formulated as a multi-objective optimisation problem, whereby the annual internal (investment plus operational) costs and external costs are simultaneously minimised. The external costs can serve as a balancing approach for emission reduction. Furthermore, it can encourage the reduction of the overall environmental impact beyond the regulatory emission constraints. A genetic optimisation algorithm is integrated into the selection tool, which optimises the combinations of abatement options that are subject to the applicable constraints. The functioning of the methodology is evaluated by case studies for the NAVAIS subjects: a battery-electric road ferry and a diesel-electric workboat, both for European waters. The results can provide useful insights into the concept design space of feasible combinations that comply with emission regulations.","Emission reduction; Multi-objective optimisation; Decision support; External costs; Early design stage","en","master thesis","","","","","","","","","","","","Marine Technology","",""
"uuid:17c90fbb-dea3-43b5-8ea7-08eda664cf5d","http://resolver.tudelft.nl/uuid:17c90fbb-dea3-43b5-8ea7-08eda664cf5d","An LT- ready and economically feasible renovation façade design","Kounaki, Stamatia (TU Delft Architecture and the Built Environment; TU Delft Architectural Engineering +Technology)","Konstantinou, Thaleia (mentor); van den Ham, Eric (graduation committee); Stellingwerff, Martijn (graduation committee); Delft University of Technology (degree granting institution)","2019","The refurbishment of the Dutch post- war residential building stock provides a great potential for energy and CO2 emissions savings, since a large number of buildings were built before 1975 and under low or no energy standards. However, their refurbishment and especially zero-on-the-meter, is too expensive and it is not possible to find the required funding. Following the guidelines of the Project MVI-Energie, which proposes disconnecting from the gas and decreasing the heating demand (that can reach up to 70% of energy demand) by insulating as such to be able to change to low temperature heating. This study focused on portiekapartments and proposed a refurbishment strategy to achieve LT-ready, while lowering the costs, time and nuisance to the residents. The facade criteria regarding the energy efficiency are the results of a parametric simulation, which includes every significant facade variable and the innovative process of applying external insulation at the optimal percentage to achieve LT-ready. The results of this simulation - diagram of solutions- can be used to upscale the research and as a guideline to quickly and easily understand if this strategy can be applied to other buildings and what it will need, based on their facade characteristics. For the technical solution a prefabricated small size insulating panel is used, that reinforces the economically feasible and fast refurbishment idea. An important aspect of the solution is to be adaptive (for example provide different cladding options) and can be part of a two-step refurbishment by providing the opportunity in the future to be continued to deeper refurbishment if needed.","refurbishment strategy; dutch post-war portiekapartments; facade insulation; LT-ready; economically feasible refurbishment; prefabricated panel; parametric simulations; Multi Objective Optimisation","en","master thesis","","","","","","","","","","","","Architecture, Urbanism and Building Sciences | Building Technology | Sustainable Design","",""
"uuid:6ead1611-990d-40a0-983e-235535c84ce0","http://resolver.tudelft.nl/uuid:6ead1611-990d-40a0-983e-235535c84ce0","Design Optimisation: Computationally Optimised Bridge Design","Nikolić, Momir (TU Delft Architecture and the Built Environment)","Delft University of Technology (degree granting institution)","2019","The computational optimisation process have been used in the structural design for long time now. Even though it occasionally appears in architectural practice, it has not yet been completely developed and its full potential has not yet been recognised.
In order to further explore the potential of computational optimisation in architectural design process, and to establish a well performing work-flow, this thesis researches on use of it, on the case study of a complex bridge design.
Suitable design assignment has been recognised in the south-east area of Rotterdam, with a large crossing over Nieuwe Maas river. Due to complex context and many different groups of stakeholders in this area, this bridge design assignment represents a suitable case study for development and testing of aforementioned design process.
Already established parametric design work-flow serves as a base of this project, while the exploratory spirit of it is recognised in use of architectural parameters and objectives for optimisation, as an improvement on traditional, only structural optimisation. Modern optimisation software and newly developed algorithms are used in order to process enlarged pool of solutions created by introducing larger number of parameters and objectives.
As a result, this project aims to develop a well integrated design proposal for a bridge over Nieuwe Maas. In that way, it explores the potential of using computational optimisation within the architectural design process as it hopes to create a stepping stone based on which further development of this design work-flow can be conducted.","Bridge Design; Multi Objective Optimisation; Architectural Optimisation; Optimisation Algorithms","en","master thesis","","","","","","","","","","","","Architecture, Urbanism and Building Sciences","",""
"uuid:58c7e8af-f9fd-44e7-afab-621027b41422","http://resolver.tudelft.nl/uuid:58c7e8af-f9fd-44e7-afab-621027b41422","A Multi-Objective Approach for the Analysis of a Water-Food-Ecosystems Nexus at Basin Scale","Farrokhzadeh, S. (TU Delft Water Resources; University of Sistan and Baluchestan); Abraham, E. (TU Delft Water Resources); Ertsen, M.W. (TU Delft Water Resources)","","2019","","Water-Food-Ecosystems Nexu; Water Allocation; Cropping Pattern; Multi-objective Optimisation; Uncertainty Analysis","en","poster","","","","","","","","","","","Water Resources","","",""
"uuid:d663c13f-ccd6-49b3-9445-ec06c8052618","http://resolver.tudelft.nl/uuid:d663c13f-ccd6-49b3-9445-ec06c8052618","Investigating trade-offs between the operating cost and green house gas emissions from water distribution systems","Menke, Ruben (Imperial College London); Kadehjian, K (Imperial College London); Abraham, E. (TU Delft Water Resources); Stoianov, Ivan (Imperial College London)","","2017","For electricity grids with an increasing share of intermittent renewables, the power generation mix can have significant daily variations. This leads to time-dependent emission intensities and volatile electricity prices in the day-ahead and spot market tariffs that can be better utilised by energy intensive industries such as water supply utilities. A multi-objective optimisation method for scheduling the operation of pumps is investigated in this paper for the reduction of both electricity costs and greenhouse gas emissions for a benchmark water distribution system. A set of energy supply scenarios has been formulated based on future projections from National Grid plc (UK) in order to investigate the range of cost savings and emission reductions that could be possibly achieved. Pump scheduling options with fixed time-of-use and day ahead market tariffs are analysed in order to compare potential reduction tradeoffs for both electricity costs and greenhouse gas emissions using Pareto optimality. The presented analysis concludes that the explicit inclusion of greenhouse gas emission reductions in optimising the scheduling of pumps operation in water distribution systems could provide considerable benefits; however, more compelling fiscal and regulatory incentives are needed.","Multi-objective optimisation; Water distribution systems; Pump scheduling","en","journal article","","","","","","","","","","","Water Resources","","",""
"uuid:3e4af7f6-9a7c-4c5b-bb55-9c1b928c9d0c","http://resolver.tudelft.nl/uuid:3e4af7f6-9a7c-4c5b-bb55-9c1b928c9d0c","Embodied Optimisation Tool for low-rise office buildings in steel","Koonath Surendran, S.","Rots, J.G. (mentor); Coenders, J.L. (mentor); Welleman, J.W. (mentor); Den Hollander, J.P. (mentor); Bonnema, B.H. (mentor); Rolvink, A. (mentor)","2014","This thesis investigates the development and application of a computational tool that optimises the conceptual stage design of a building to have minimum embodied energy and some aspects of the operating energy, depending on the adaptability required. For this purpose, a parametric computational framework for sustainable building design was developed and implemented by the tool. The working prototype of the tool focuses on low-rise rectangular grid office buildings in steel. The various competing objectives are optimised by applying multi-objective optimisation techniques. The Embodied Optimisation Tool has been developed as a plugin within Grasshopper, for 3D modeling tool Rhinoceros, in collaboration with Arup, Bouwen Met Staal and Tata Steel.","multi-objective optimisation; embodied energy; operating energy; adaptability; sustainability; parametric design; computational tool; conceptual stage","en","master thesis","","","","","","","","","Civil Engineering and Geosciences","Structural Engineering","","Structural Design","",""
"uuid:402b91f8-8b85-4c7c-9cc0-086471d5ebd8","http://resolver.tudelft.nl/uuid:402b91f8-8b85-4c7c-9cc0-086471d5ebd8","Evo-Devo in the Sky","Janssen, P.","","2013","Designers interested in applying evo-devo-design methods for performance based multi-objective design exploration have typically faced two main hurdles: its too hard and too slow. An evo-devo-design method is proposed that effectively overcomes the hurdles of skill and speed by leveraging two key technologies: computational workflows and cloud computing. In order to tackle the skills hurdle, Workflow Systems are used that allow users to define computational workflows using visual programming techniques. In order to tackle the speed hurdle, cloud computing infrastructures are used in order to allow the evolutionary process to be parallelized. We refer to the proposed method as Evo-Devo In The Sky (EDITS). This paper gives an overview of both the EDITS method and the implementation of a software environment supporting the EDITS method. Finally, a case-study is presented of the application of the EDITS method.","evolutionary algorithms; multi-objective optimisation; workflow system; cloud computing; parametric modelling","en","conference paper","","","","","","","","","","","","","",""
"uuid:3c921dd4-c9f5-439e-a412-afd0856ba689","http://resolver.tudelft.nl/uuid:3c921dd4-c9f5-439e-a412-afd0856ba689","An Integrated Approach to Aircraft Modelling and Flight Control Law Design","Looye, G.H.N.","Mulder, J.A. (promotor); Chu, Q.P. (promotor)","2008","The design of flight control laws (FCLs) for automatic and manual (augmented) control of aircraft is a complicated task. FCLs have to fulfil large amounts of performance criteria and must work reliably in all flight conditions, for all aircraft configurations, and in adverse weather conditions. Consequently, a large part of the FCL design process involves extensive simulation analyses, hardware-in-the-loop testing, and, eventually, flight testing. Multi-disciplinary aspects hereby play an important role. For example, control laws heavily influence (aerodynamic) loads on the airframe during manoeuvring and in turbulence, as well as flutter stability of the structure. These aspects are extensively addressed, but only -after- the actual design phase. As a consequence, problems that arise with other disciplines usually give rise to re-design of control law functions. This thesis proposes an FCL design process that allows multi-disciplinary aspects to be addressed from the beginning. In the first place, this requires multi-disciplinary aspects to be present in the aircraft dynamics models used for FCL design. To this end, the use of object-oriented modelling techniques is proposed, which, in contrast to contemporary techniques, inherently supports the development of models consisting of components from various engineering areas. As a specific application, its use for development of integrated flight mechanics and aeroelastic aircraft models is discussed. In the second place, multi-disciplinary FCL design requires a means to automate tuning of design parameters, since consideration of the many additional criteria make manual parameter synthesis very elaborate. For this reason, the use of multi-objective optimisation is proposed. This technique allows parameters to be optimised with respect to many, possibly conflicting, design criteria via a so-called min-max approach. The process is demonstrated on the design of control laws for automatic landing (autoland) of a passenger aircraft. The certification of autoland systems requires extensive Monte Carlo (MC) analyses to be performed in order to show that landing mishaps in all sorts of extreme conditions are very unlikely. The proposed design process allows the MC analyses to be directly addressed in the synthesis of control law parameters, so that MC analyses for certification can be passed in one shot. The proposed multi-disciplinary design process further allows the control design department to increase participation in aircraft preliminary design, by providing a means for rapid control law prototyping. This methodology allows nonlinear control laws for a specific aircraft design status to be automatically generated from an object-oriented implementation of a current flight dynamics model, using the technique of feedback linearisation and the possibility of automatic model inversion from object-oriented model implementations. For the control department, rapid prototyping allows for quick experimentation with controller structures, the selection of command variables, etc. For other engineering departments, the methodology results in early availability of representative control laws to analyse dynamic flight characteristics of the current aircraft design status.","flight dynamics; flight control; aeroelasticity; multi-disciplinary modelling; multi-objective optimisation","en","doctoral thesis","","","","","","","","","Aerospace Engineering","","","","",""
"uuid:e11b3a08-aad0-4770-83e1-98326284fcf4","http://resolver.tudelft.nl/uuid:e11b3a08-aad0-4770-83e1-98326284fcf4","Robust Design Optimisation by mode FRONTIER","Clarich, A.; Carlo, P.; Valentino, P.","","2006","This presentation deals with industrial applications, in aeronautic and turbomachinery fields, of multi-objective and robust design optimisation, through the utilisation of the multi-objective optimisation and design environment modeFrontier. This code allows the easy process integration of any CAD/CAE commercial tool, and drives the designer in a fully automatic optimisation of the objectives he has defined in his model. A particular relief is given in this presentation in the application of the Game Theory algorithm implemented in modeFrontier, MOGT, that allow to reach an optimal compromise solution between the contrasting objectives in the fewest time as possible. To reduce the number of simulations required is in fact particularly important in the design under uncertainties, or so called Robust Design Optimisation, since this typology of optimisation, that is becoming predominant in aeronautic design, usually requires an higher number of simulations than the simple deterministic optimisation, since the behaviour of each candidate solution is to be analysed for several samples characterised by fluctuations (uncertainties) of input variables. The application of MOGT algorithm and Response Surfaces, both available in modeFrontier, will prove their efficiency in the Design under Uncertainties cases proposed in this presentation.","Robust Design Optimisation; multi-objective optimisation; response surfaces; Game Theory Optimisation algorithms; aeronautics; turbomachinery","en","conference paper","","","","","","","","","","","","","",""