Machine learning techniques promise to greatly accelerate structural design and optimization. In this thesis, deep learning and active learning techniques are applied to different non-convex structural optimization problems. Finite Element Analysis (FEA) based standard optimization methods for aircraft panels with bio-inspired curvilinear stiffeners are computationally expensive. The main reason for employing many of these standard optimization methods is the ease of their integration with FEA. However, each optimization requires multiple computationally expensive FEA evaluations, making their use impractical at times. To accelerate optimization, the use of Deep Neural Networks (DNNs) is proposed to approximate the FEA buckling response. The results show that DNNs obtained an accuracy of 95% for evaluating the buckling load. The DNN accelerated the optimization by a factor of nearly 200. The presented work demonstrates the potential of DNN-based machine learning algorithms for accelerating the optimization of bio-inspired curvilinearly stiffened panels. But, the approach could have disadvantages for being only specific to similar structural design problems, and requiring large datasets for DNNs training. An adaptive machine learning technique called active learning is used in this thesis to accelerate the evolutionary optimization of complex structures. The active learner helps the Genetic Algorithms (GA) by predicting if the possible design is going to satisfy the required constraints or not. The approach does not need a trained surrogate model prior to the optimization. The active learner adaptively improve its own accuracy during the optimization for saving the required number of FEA evaluations. The results show that the approach has the potential to reduce the total required FEA evaluations by more than 50%. Lastly, the machine learning is used to make recommendations for modeling choices while analyzing a structure using FEA. The decisions about the selection of appropriate modeling techniques are usually based on an analyst's judgement based upon their knowledge and intuition from past experience. The machine learning-based approach provides recommendations within seconds, thus, saving significant computational resources for making accurate design choices. / Doctor of Philosophy / This thesis presents an innovative application of artificial intelligence (AI) techniques for designing aircraft structures. An important objective for the aerospace industry is to design robust and fuel-efficient aerospace structures. The state of the art research in the literature shows that the structure of aircraft in future could mimic organic cellular structure. However, the design of these new panels with arbitrary structures is computationally expensive. For instance, applying standard optimization methods currently being applied to aerospace structures to design an aircraft, can take anywhere from a few days to months. The presented research demonstrates the potential of AI for accelerating the optimization of an aircraft structures. This will provide an efficient way for aircraft designers to design futuristic fuel-efficient aircraft which will have positive impact on the environment and the world.

A new noise metric has been developed that may be used for optimization problems involving aerodynamic noise from a clean wing. The modeling approach uses a classical trailing edge noise theory as the starting point. The final form of the noise metric includes characteristic velocity and length scales that are obtained from three-dimensional, steady, RANS simulations with a two- equation k-omega turbulence model. The noise metric is not the absolute value of the noise intensity, but an accurate relative noise measure as shown in the validation studies. One of the unique features of the new noise metric is the modeling of the length scale, which is directly related to the turbulent structure of the flow at the trailing edge. The proposed noise metric model has been formulated so that it can capture the effect of different design variables on the clean wing airframe noise such as the aircraft speed, lift coefficient, and wing geometry. It can also capture three-dimensional effects which become important at high lift coefficients, since the characteristic velocity and the length scales are allowed to vary along the span of the wing. Noise metric validation was performed with seven test cases that were selected from a two-dimensional NACA 0012 experimental database. The agreement between the experiment and the predictions obtained with the new noise metric was very good at various speeds, angles of attack, and Reynolds Number, which showed that the noise metric is capable of capturing the variations in the trailing edge noise as a relative noise measure when different flow conditions and parameters are changed. Parametric studies were performed to investigate the effect of different design variables on the noise metric. Two-dimensional parametric studies were done using two symmetric NACA four-digit airfoils (NACA 0012 and NACA 0009) and two supercritical (SC(2)-0710 and SC(2)-0714) airfoils. The three-dimensional studies were performed with two versions of a conventional transport wing at realistic approach conditions. The twist distribution of the baseline wing was changed to obtain a modified wing which was used to investigate the effect of the twist on the trailing edge noise. An example study with NACA 0012 and NACA 0009 airfoils demonstrated a reduction in the trailing edge noise by decreasing the thickness ratio and the lift coefficient, while increasing the chord length to keep the same lift at a constant speed. Both two- and three-dimensional studies demonstrated that the trailing edge noise remains almost constant at low lift coefficients and gets larger at higher lift values. The increase in the noise metric can be dramatic when there is separation on the wing. Three-dimensional effects observed in the wing cases indicate the importance of calculating the noise metric with a characteristic velocity and length scale that vary along the span. The twist change does not have a significant effect on the noise at low lift coefficients, however it may give significant noise reduction at higher lift values. The results obtained in this study show the importance of the lift coefficient on the airframe noise of a clean wing and favors having a larger wing area to reduce the lift coefficient for minimizing the noise. The results also point to the fact that the noise reduction studies should be performed in a multidisciplinary design and optimization framework, since many of the parameters that change the trailing edge noise also affect the other aircraft design requirements. It's hoped that the noise metric developed here can aid in such multidisciplinary design and optimization studies. / Ph. D.

The acoustic emission technique is an effective way to acquire crack information from material bodies at the microscopic level. Monitoring of the acoustic emission events provides a deeper understanding regarding the structural health status of critical constructions such as bridges, railways, pipelines, pressure vessels, etc. Thanks to the acoustic emission monitoring systems, it is possible to avoid catastrophic events and save lives, time, and money. For this reason, efforts to develop new acoustic emission sensor technologies, as well as the use of current acoustic emission sensors in new research fields, will contribute to the limited literature sources. Optical sensing systems provide good alternatives to the existing sensing technologies because of their wide range of detection bandwidths, adaptation to harsh environments, and low sensitivity to electromagnetic interference. For this reason, the first part of this thesis demonstrates an optoacoustic sensing methodology that enables the detection of acoustic emissions by optics. This sensing system consists of thin-film optical filters (TFOF) and an elastic microcavity layer. The sensing mechanism is similar to the Fabry Perot structures and it relies on resonance shifts of the cavity when there is a change in the cavity thickness similar to the Fabry Perot structures. Thus, the design, fabrication, and demonstration steps of a Fabry Perot elastic microcavity have been presented. Throughout the fabrication efforts, a new deposition protocol was developed. This deposition technique has enabled the deposition of TFOF on flexible substrates via the RF-sputtering technique. Thus, a new sensing configuration has been developed using flexible optical components. In the second chapter, an optical sensing methodology based on tunable spectral filters and flexible optical components is introduced. The design, fabrication, realization, and characterization of a proof-of-concept optomechanical sensor have been presented. The design step includes optical, mechanical, and optoacoustic correlation simulations using the Transfer Matrix Method, finite element analysis, and analytical models. Moreover, the fabrication part includes multilayer deposition on silica and flexible substrates using the RF-Sputtering technique and integration of these optical components into a 3D-printed housing together with electronic components. Eventually, the performance evaluation of the optomechanical sensor has been carried out and the experimental results showed that the sensor resonance frequency is around 515 Hz and the sensor is capable of detecting static loadings from 50 Pa to 235 Pa values. In the fourth chapter, seismic vulnerability analysis of a coupled Tank-Piping System has been performed using traditional acoustic emission sensors. Real-time performance evaluation of the pipeline as well as the structural health status of the critical parts were monitored. As a result, deformation levels of each critical part were investigated, and the processing of acoustic emission signals provided a more in-depth view of damage level of the analyzed components. Throughout the thesis, TFOFs are an integral part of this thesis. Therefore, both the design and simulation of TFOFs play a crucial role throughout this research work. The Transfer Matrix Method is used to simulate the optical performance of TFOFs. Moreover, in the final chapter, an automated design framework is presented for the design of TFOFs using a nature-inspired machine learning approach called Genetic algorithm. This design approach enables the design of sophisticated geometric configurations with unique optical capabilities. Therefore, not only the improvement of sensor response but also the new ways in the development of novel optical systems are demonstrated in this final chapter.

<p> This thesis proposes a sensitivity solver for frequency-domain electromagnetic (EM) simulators based on volume methods such as the finite-element method (FEM). The proposed sensitivity solver computes S-parameter Jacobians directly from the field solutions available from the EM simulation. It exploits the computational efficiency of the self-adjoint sensitivity analysis (SASA) approach where only one EM simulation suffices to obtain both the responses and their gradients in the designable parameter space. The proposed sensitivity solver adopts the system equations of the finite-difference frequency-domain (FDFD) method.</p> <p> There are three major advantages to this development: (1) the Jacobian computation is completely independent of the simulation engine, its grid and its system equations; (2) the implementation is straightforward and in the form of a post-processing algorithm operating on the exported field solutions; and (3) it is computationally very efficient-time requirements are negligible in comparison with conventional field-based optimization procedures utilizing Jacobians computed via response-level finite differences or parameter sweeps.</p> <p> The accuracy and the efficiency of the proposed sensitivity solver are verified in the sensitivity analysis and the gradient-based optimization of filters and antennas. Compared to the finite-difference approximation, drastic reduction of the time required by the overall optimization process is achieved. All examples use a commercial finite-element simulator.</p> <p> Suggestions for future research are provided.</p> / Thesis / Master of Applied Science (MASc)

Today, many embedded or cyber-physical systems, e.g., in the automotive domain, comprise several control applications, sharing the same platform. It is well known that such resource sharing leads to complex temporal behaviors that degrades the quality of control, and more importantly, may even jeopardize stability in the worst case, if not properly taken into account. In this thesis, we consider embedded control or cyber-physical systems, where several control applications share the same processing unit. The focus is on the control-scheduling co-design problem, where the controller and scheduling parameters are jointly optimized. The fundamental difference between control applications and traditional embedded applications motivates the need for novel methodologies for the design and optimization of embedded control systems. This thesis is one more step towards correct design and optimization of embedded control systems. Offline and online methodologies for embedded control systems are covered in this thesis. The importance of considering both the expected control performance and stability is discussed and a control-scheduling co-design methodology is proposed to optimize control performance while guaranteeing stability. Orthogonal to this, bandwidth-efficient stabilizing control servers are proposed, which support compositionality, isolation, and resource-efficiency in design and co-design. Finally, we extend the scope of the proposed approach to non-periodic control schemes and address the challenges in sharing the platform with self-triggered controllers. In addition to offline methodologies, a novel online scheduling policy to stabilize control applications is proposed.

L’étude des réseaux de microcapteurs sans fil met clairement en évidence la contrainte principale de l’autonomie en énergie. En effet, ces microcomposants autonomes et communicants appelés aussi nœuds du réseau sont dispersés dans des lieux parfois peu ou pas accessibles. L’objectif de notre travail est de proposer une méthode de conception d’un émetteur-récepteur adapté à ce type de réseaux. Partant d’une modélisation au niveau système mettant en relief la part prépondérante du module radiofréquence sur la consommation moyenne d’un nœud, nous avons développé trois nouveaux outils de conception correspondant à différents niveaux de modélisation de la chaine de communication. Leur utilisation conjointe et les résultats de simulations obtenus nous offrent la possibilité de mettre en relation les spécifications et les performances d’un module radiofréquence avec la consommation. L’association de ces outils dans une méthode de conception itérative nous a permis de dimensionner une chaine de communication en fonction d’une contrainte de consommation. Finalement, nous avons conçu, fabriqué et testé, un amplificateur faible bruit (LNA ou Low Noise Amplifier) à 868 MHz qui présente des caractéristiques très intéressantes en termes de consommation. / Survey on wireless microsensor networks highlights the main constraint of energy autonomy. In fact, these autonomous and communicating microcomponents named network nodes are scattered into few or not open environment. The goal of our work is to propose a transceiver design method adapted to microsensor networks. After a demonstration of predominant part of RF into the mean power consumption of a microsensor node, we developed three new simulation tools which correspond to different level of transceiver modelling. Their use and obtained simulation results demonstrate the relation between transceiver specifications and performances with power consumption. The association of these tools was used to propose a new design method under power consumption constraint. Finally, we designed, produced and tested a 868 MHz Low Noise Amplifier which presents interesting power consumption characteristics.

Absorvedores de vibração são comumente usados em aplicações com intuito de reduzir indesejadas amplitudes de vibração de estruturas e maquinas vibrantes. O conceito de um absorvedor de vibração linear consiste na ideia de projetar um subsistema com frequência de ressonância coincidente com uma dada frequência de interesse, tal que a amplitude de vibração do sistema primário e significativamente reduzida quando comparada a situação original, sem o absorvedor de vibração. Porem, uma deficiência dos absorvedores de vibração lineares típicos e sua estreita faixa de frequência de operação. Para superar essa deficiência, muitas tentativas de solução usando subsistemas não lineares tem sido propostas na literatura, ja que se apropriadamente projetados, eles podem aumentar a faixa de frequência de absorção de vibração e/ou melhorar a redução das amplitudes de vibração do sistema primário. Contudo, a síntese e o projeto de tais absorvedores não lineares não e tão simples e direta como no caso linear. Baseado na geometria de uma topologia proposta e encontrada na literatura, que compreende a inclusão de uma montagem do tipo snap through truss no lugar da mola linear do absorvedor de vibração, este trabalho tem intenção de apresentar um estudo sobre o projeto e otimização de um absorvedor dinâmico de vibrações não linear. Portanto, o efeito dos parâmetros do absorvedor e analisado quanto as perspectivas de redução das amplitudes de vibração do sistema principal como também de aumento da faixa de frequência de operação. A analise paramétrica do absorvedor foi promovida para responder questões sobre as variáveis de projeto, tanto físicas como geométricas. Realizou-se otimização do absorvedor com objetivo de sintoniza-lo a frequência de trabalho desejada, através de busca extensiva e algoritmos genéticos. Os resultados mostram que o absorvedor não linear proposto pode ser mais efetivo que seu correspondente linear em ambos os aspectos, na redução da máxima amplitude de vibração e no aumento da faixa de frequência de absorção. Portanto, apesar da dificuldade inicial de projeto, esse tipo de absorvedor representa uma alternativa interessante na atenuação das amplitudes de vibração ao longo de uma extensa faixa de frequência. / Dynamic vibration absorbers are commonly used in several applications in order to reduce undesired vibration amplitudes of vibrating machinery and structures. The concept of a linear vibration absorber is based on the idea of designing a subsystem with a resonance frequency coincident with a given frequency of interest such that the vibration amplitude of the primary system is significantly reduced when compared to the original situation (without the vibration absorber). But one of the known handicaps of typical linear vibration absorbers is their narrow frequency range of operation. To overcome this handicap, a number of tentative solutions have been proposed in the literature using nonlinear subsystems. If properly designed, they could enlarge the frequency range of vibration absorption and/or improve vibration reduction of the primary system. However, the synthesis and design of such nonlinear absorbers are not as straightforward as for their linear counterpart. A proposed design found in the open literature consists of replacing the linear spring of the vibration absorber by a nonlinear snap-through truss. This work aims to present a study on the design and optimization of a nonlinear dynamic vibration absorber based on snap-through absorber geometry. The effect of the absorber parameters was analyzed on both, the primary system vibration amplitude reduction and the frequency range of operation. Parametric analyses of the absorber were carried out to answer questions about the physical and geometric design variables. The absorber optimization was performed in two different ways, by extensive search and genetic algorithms, in order to tune it in the desired working frequency. The results show that the proposed nonlinear vibration absorber may be more effective than its linear counterpart both in terms of maximum vibration amplitude reduction and absorption frequency-range. Therefore, despite the increased design complexities such an absorber is an interesting alterna- tive in attenuating vibration amplitudes over a wide frequency range.

Layout optimization of offshore wind farms seeks to automate the design of the wind farm and the placement of wind turbines such that the proposed wind farm maximizes its potential. The optimization of an offshore wind farm layout therefore seeks to minimize the costs of the wind farm while maximizing the energy extraction while considering the effects of wakes on the resource; the electrical infrastructure required to collect the energy generated; the cost variation across the site; and all technical and consenting constraints that the wind farm developer must adhere to. As wakes, electrical losses, and costs are non-linear, this produces a complex optimization problem. This thesis describes the design, development, validation, and initial application of a new framework for the optimization of offshore wind farm layouts using either a genetic algorithm or a particle swarm optimizer. The developed methodology and analysis tool have been developed such that individual components can either be used to analyze a particular wind farm layout or used in conjunction with the optimization algorithms to design and optimize wind farm layouts. To accomplish this, separate modules have been developed and validated for the design and optimization of the necessary electrical infrastructure, the assessment of the energy production considering energy losses, and the estimation of the project costs. By including site-dependent parameters and project specific constraints, the framework is capable of exploring the influence the wind farm layout has on the levelized cost of energy of the project. Deploying the integrated framework using two common engineering metaheuristic algorithms to hypothetical, existing, and future wind farms highlights the advantages of this holistic layout optimization framework over the industry standard approaches commonly deployed in offshore wind farm design leading to a reduction in LCOE. Application of the tool to a UK Round 3 site recently under development has also highlighted how the use of this tool can aid in the development of future regulations by considering various constraints on the placement of wind turbines within the site and exploring how these impact the levelized cost of energy.

Acknowledging the goal of reduced aircraft weight, there is a need to improve on conservative design techniques used in industry. Minimisation of laminate in-plane elastic energy is used as an appropriate in-plane performance marker to assess the weight saving potential of new design techniques. MATLAB optimisations using a genetic algorithm were used to find the optimal laminate variables for minimum in-plane elastic energy and/or damage tolerance for all possible loadings. The use of non-standard angles was able to offer equivalent, if not better in-plane performance than standard angles, and are shown to be useful to improve the ease of manufacture. Any standard angle laminate stiffness was shown to be able to be matched by a range of two non-standard angle ply designs. This non-uniqueness of designs was explored. Balancing of plus and minus plies about the principal loading axes instead of themanufacturing axes was shown to offer considerable potential for weight saving as the stiffness is better aligned to the load. Designing directly for an uncertain design load showed little benefit over the 10% ply percentage rule in maintaining in-plane performance. This showed the current rule may do a sufficient job to allow robustness in laminate performance. This technique is seen useful for non-standard angle design that lacks an equivalent 10% rule. Current use of conservative damage tolerance strain limits for design has revealed the need for more accurate prediction of damage propagation. Damage tolerance modelling was carried out using fracture mechanics for a multi-axial loading considering the full 2D strain energy and improving on current uni-axial models. The non-conservativeness of the model was evidenced to be from assumptions of zero post-buckled stiffness. Preliminary work on conservative multi-axial damage tolerance design, independent of thickness, is yet to be confirmed by experiments.

The need of water and the limited sources, force the researchers to find the most economical and feasible solution in the design of a water distribution network. In this study, reliability and optimization of a water distribution network are taken into account together in the design stage of the network. The relationship between reliability of a water distribution network and its cost is examined during the design of a water distribution network. A methodology for deciding the reliability level of the selected design is proposed by examining the reliability-cost relationship. The design alternatives for the case study area are obtained by the aid of a commercially available software WADISO employing partial enumeration optimization technique. The reliability value for each of the design alternative is calculated according to Misirdali (2003)&rsquo / s adaptation based on the methodology proposed by Bao and Mays (1990) by the aid of a hydraulic network solver program HapMam prepared by Nohut&ccedil / u (2002). For purposes of illustration, the skeletonized form of Ankara Water Distribution Network subpressure zone (N8-1) is taken as the case study area. The methodology in this study, covering the relation between the reliability and the cost of a water distribution network and the proposed reliability level can be used in the design of new systems.