Despite significant advancements in the understanding of fluid flows, combustion and material technologies, hypersonic flight still presents numerous technological challenges. In hypersonic vehicles turbulence is critical in controlling heat generation in the boundary layer, mixing inside the combustor, generation of acoustic noise, and mass flow in the intake. The study of turbulence in highly compressible flows is challenging compared to incompressible due to a drastic change in the behavior of pressure and a relaxation of the incompressibility constraint. In addition fluid flow inside a flight vehicle is complicated by wall-effects, heat generation and complex boundary conditions. Homogeneous shear flow contains most of the relevant physics of boundary and mixing layers without the aforementioned complicating effects. In this work we aim to understand and characterize the role of pressure, velocity-pressure interaction, velocity-thermodynamics interaction in the late-stage transition-to-turbulence regime in a high speed shear dominated flow by studying the evolution of perturbations in in a high Mach number homogeneous shear flow. We use a modal-analysis based approach towards understanding the statistical behavior of turbulence. Individual Fourier waves constituting the initial flow field are studied in isolation and in combination to understand collective statistical behavior. We demonstrate proof of concept of novel acoustic based strategies for controlling the onset of turbulence. Towards this goal we perform direct numerical simulations (DNS) in three studies: (a) development and evaluation of gas kinetic based numerical tool for DNS of compressible turbulence, and perform detailed evaluation of the efficacy of different interpolation schemes in capturing solenoidal and dilatational quantities, (b) modal investigation in the behavior of pressure and isolation of linear, non-linear, inertial and pressure actions, and (c) modal investigation in the possible acoustic based control strategies in homogeneously sheared compressible flows. The findings help to understand the manifestation of the effects of compressibility on transition and turbulence via the velocity-pressure interactions and the action of individual waves. The present study helps towards the design of control mechanisms for compressible turbulence and the development of physically consistent pressure strain correlation models.

This thesis is aimed at identifying the noise sources of a wheeled tractor to reduce the noise levels below the legislative limits by controlling noise sources through proper methodologies. The study focuses firstly on identifying the noise sources of a wheeled tractor by using proper noise source identification techniques. These techniques can be summarized as sound intensity measurements, sound power level determination studies and spectral analysis of the noise data acquired in the tests. Simple sound intensity mapping techniques are used and the intensity contour maps are generated to identify the noise sources. Most important and effective noise sources are identified and the critical noise sources are focused to apply appropriate noise control strategies not only at the prototype production stages but also at the early design stages. Consequently, upon consideration of both structure-borne and flow-induced noise, the pass-by noise level and the operator&rsquo / s ear noise levels of the tractor are reduced by nearly 3 dB (A) through application of proper noise control strategies.

This thesis (1) reports a new Dynamic Programming (DP) approach, and (2) reports a Real Time Control strategy to optimize the energy management of a Hybrid Electric Vehicle(HEV). Increasing environmental concerns and rise in fuel prices in recent years has escalated interest in fuel efficient vehicles from government, consumers and car manufacturers. Due to this, Hybrid electric vehicles (HEV) have gained popularity in recent years. HEV’s have two degrees of freedom for energy flow controls, and hence the performance of a HEV is strongly dependent on the control of the power split between thermal and electrical power sources. In this thesis backward-looking and forward-looking control strategies for two HEV architectures namely series and parallel HEV are developed. The new DP approach, in which the state variable is not discretized, is first introduced and a theoretical base is established. We then prove that the proposed DP produces globally optimal solution for a class of discrete systems. Then it is applied to optimize the fuel economy of HEV's. Simulations for the parallel and series HEV are then performed for multiple drive cycles and the improved fuel economy obtained by the new DP is compared to existing DP approaches. The results are then studied in detail and further improvements are suggested. A new Real Time Control Strategy (RTCS) based on the concept of preview control for online implementation is also developed in this thesis. It is then compared to an existing Equivalent Cost Minimization Strategy (ECMS) which does not require data to be known apriori. The improved fuel economy results of the RTCS for the series and parallel HEV are obtained for standard drive cycles and compared with the ECMS results

Τα «Συστήματα Αυτόματης Απόδειξης Θεωρημάτων-ΣΑΑΘ» (Automatic Theorem Proving Systems-ATP Systems) είναι συστήματα βασισμένα στη λογική πρώτης τάξεως, τα οποία μπορούν από ένα σύνολο λογικών προτάσεων να συνάγουν την αλήθεια μιας δεδομένης λογικής πρότασης με αυτόματο τρόπο. Η διαδικασία της απόδειξης στα περισσότερα ΣΑΑΘ στηρίζεται στην αρχή της επίλυσης, τον ισχυρότερο κανόνα λογικής εξαγωγής συμπερασμάτων, και την αντίφαση της επίλυσης, μια διαδικασία που εξασφαλίζει την ορθότητα των συμπερασμάτων. Ο ACT-P είναι ένα ΣΑΑΘ που στηρίζεται στην αρχή της επίλυσης και την αντίφαση της επίλυσης, γραμμένο στο εργαλείο GCLISP Developer 5.0 της Gold-Hill, και διαθέτει μια βιβλιοθήκη γνωστών στρατηγικών ελέγχου της διαδικασίας απόδειξης, προσφέροντας τη δυνατότητα στον χρήστη να ορίσει κάθε φορά ένα (κατάλληλο) συνδυασμό στρατηγικών. Στην εργασία αυτή έγινε κατ’ αρχήν μεταφορά του ACT-P σε LispWorks, ένα δυναμικότερο εργαλείο ανάπτυξης εφαρμογών σε Lisp. Επιπλέον, ο χρήστης μέσω του νέου παραθυρικού περιβάλλοντος μπορεί να βλέπει δυο διαφορετικές λύσεις του ίδιου προβλήματος, τη συνοπτική και αναλυτική λύση. Στη συνέχεια, έγινε έλεγχος της καλής λειτουργίας του ACT-P και των στρατηγικών του μέσω δοκιμών με προβλήματα που προέρχονται από την TPTP (Thousands of Problems for Theorem Provers), μια γνωστή βιβλιοθήκη προβλημάτων για ΣΣΑΘ συστήματα στο Διαδίκτυο, και έγιναν οι απαραίτητες διορθώσεις έτσι ώστε να επιλύει προβλήματα από διάφορες κατηγορίες προβλημάτων της βιβλιοθήκης TPTP. Τέλος, έγινε μια μελέτη χρήσης διαφόρων συνδυασμών στρατηγικών ελέγχου για διάφορα προβλήματα της TPTP και εξήχθησαν χρήσιμα συμπεράσματα για την καταλληλότητά τους και την αποδοτικότητά τους σε σχέση με το είδος των προβλημάτων. / Automatic Theorem Proving Systems (ATP Systems) are based on First Order Logic (FOL) and are able to automatically prove the truth of logical sentence. The proof procedure in most ATP Systems uses the resolution principle which is the strongest existing inference rule, and the resolution refutation process which ensure soundeness of the conclusion. The ACT-P is an ATP System which uses the resolution principle and the resolution refutation and it is written in GCLISP Developer 5.0 of Gold-Hill. ACT-P has a library of strategies to control the proof process, and gives users the ability to assign to specify a suitable combination of strategies. In this dissertation a new window based interface is developed for ACTP in Lispworks, which is a powerful tool for developing Lisp applications. The interface gives to the user a more thorough view of the solving process. Moreover, the user can see two different solutions of the problem, the brief and the analytic one. In addition, the functionality and the strategies of ACTP were tested on problems from the TPTP (Thousands of Problems for Theorem Provers) which is a known library of problems for ATP Systems on the web. ACTP has been improved so as to solve problems from various categories of the TPTP library. Finally, different strategy combinations for solving problems from various categories of TPTP library were studied, leading to useful conclusions about the suitability and the performance of the different combinations depending on the problems.

The automotive industry is currently undergoing a paradigm shift. One such example in the next generation steering is the Steer-by-Wire (SbW) technology. SbW comes with a lot of advantages but one of the big challenges is to provide the driver with a realistic steering feel. More precisely, steering feel can be defined as the relationships between the steering wheel torque, the steering wheel angle and the dynamics of the vehicle. Accordingly, the first contribution of this work will be to present transfer functions between these quantities that resemble those observed in traditional steering systems. The steering feel/feedback is then achieved by an electric motor which can be controlled by different control strategies. In this thesis three different control strategies are investigated. The first straightforward strategy is called open loop since there is no feedback controller in the system. The second strategy is torque feedback control and the third strategy is angle feedback control. All three systems are evaluated in terms of reference tracking, stability, robustness and sensitivity. Here reference tracking is defined as tracking a desired transfer function. The desired transfer function is denoted as the reference generator. When fulfilling the requirements the analysis shows that the torque feedback system has a better reference tracking than the other evaluated systems. It is also concluded that the open loop system has a compromised reference tracking compared to the torque and angle feedback systems. Since the SbW technology is still an undergoing area of research within the automotive sector this work can be used as a basis for choice of control strategy for steering feedback systems and also as a guideline for future hardware choices.

Human African Trypanosomiasis (HAT) is a vector-borne disease transmitted by the bite of the tsetse fly that results in high human morbidity and mortality. The propagation of the disease has been linked to environmental factors, and understanding the vector’s habitat is vital to its control. There is no HAT vaccine, but biological control of the vector has been successful in reducing HAT incidence. However, in recent years the disease has re-emerged and spread. Due to insufficient knowledge of HAT endemic foci, the disease management remains challenging. To achieve effective deployment of control strategies, accurate knowledge of the spatial distribution of the HAT vector is vital. The current study is based in Nigeria, and looks at part of Delta State, and a part of Jigawa State, in which HAT has been identified. The work utilizes remote sensing satellite imaging and fuzzy logic to develop a HAT vector habitat classification scheme, to explore the dynamics of HAT propagation. The goal was to develop a surveillance methodology to identify factors that influence HAT epidemiology. Land cover and ancillary data were integrated to classify HAT vector habitat using geospatial-fuzzy multicriteria analysis. The work highlights the significance of geospatial techniques where epidemiological data are limited, for improving understanding of HAT. This study helped distinguish HAT vector habitat into different zones (breed, feed and rest), which allowed the direction and magnitude of HAT, a n d factors influencing propagation to be determined. This helped identify ‘HAT priority intervention areas’. The study findings suggested propagation of HAT resulted from suitability of water bodies, shrub and less-dense forest for the HAT vector, and continued exposure of human populations to these land cover classes. Overlapping of HAT vector habitat zones within built-up areas was also a cause. The study also found that HAT propagation was multidirectional, and that this may have been influenced by landscape characteristics. This novel approach can also be used in other part of Nigeria as well as adapted to investigate other diseases. In conclusion, the HAT vector habitat classification scheme is a transparent tool for policy makers for identifying vulnerable and at risk areas.

Der vorliegende Beitrag soll einen bewertenden Überblick über ausgewählte energieoptimale Regelungsstrategien für Asynchronmaschinen (ASM) geben . Diese werden zunächst in einer Simulationsumgebung implementiert und parametrisiert. Es erfolgt die Untersuchung beim Abfahren eines Fahrzyklus, sowie an einzelnen Lastpunkten. Die Resultate werden anhand ihrer Robustheit, der Praktikabilität des Regelverfahrens und nicht zuletzt am Energiesparpotential bewertet. Im ersten Teil wird die Struktur sowie Modellierung der zum Test der Regelungsalgorithmen verwendeten Komponenten vorgestellt. Auch werden die Regelungsstrategien an sich in Struktur und Funktion näher beleuchtet. Die Parametrisierung der entsprechenden Modelle erfolgt aus realen Prüfstandsdaten. Im zweiten Teil wird näher auf die zur späteren Bewertung verwendeten Verlustleistungsberechnung eingegangen. Der dritte Teil beschäftigt sich mit der Darstellung des zur Simulation verwendeten Lastzyklus. Zuletzt werden die Ergebnisse der Simulation vorgestellt und bewertet. / The present article is intended to provide an evaluative overview of four selected energy-optimized control strategies for asynchronous machines (ASM). These are first implemented and parameterized in a simulation environment. The investigation is carried out when running a driving cycle and at individual load points. The results are assessed on the basis of their robustness, practicability and, last but not least, their energy saving potential. In the first part, the structure and modeling of the components used to test the control algorithms is presented. The control strategies themselves are also examined in more detail in terms of structure and function. The corresponding models are parameterized from real test bench data. The second part deals in more detail with the power loss calculation used for the subsequent evaluation. The third deals with the representation of the load cycle used for the simulation. Finally, the results of the simulation are presented and evaluated.

Deterministic models for the transmission dynamics of HIV/AIDS and trichomonas vaginalis (TV) in a human population are formulated and analysed. The models which assumed standard incidence formulations are shown to have globally asymptotically stable (GAS) disease-free equilibria whenever their associated reproduction number is less than unity. Furthermore, both models possess a unique endemic equilibrium that is GAS whenever the associated reproduction number is greater than unity. An extended model for the co-infection of TV and HIV in a human population is also designed and rigorously analysed. The model is shown to exhibit the phenomenon of backward bifurcation, where a stable disease-free equilibrium (DFE) co-exists with a stable endemic equilibrium whenever the associated reproduction number is less than unity. This phenomenon can be removed by assuming that the co-infection of individuals with HIV and TV is negligible. Furthermore, in the absence of co-infection, the DFE of the model is shown to be GAS whenever the associated reproduction number is less than unity. This study identifies a sufficient condition for the emergence of backward bifurcation in the model, namely TV-HIV co-infection. The endemic equilibrium point is shown to be GAS (for a special case) when the associated reproduction number is greater than unity. Numerical simulations of the model, using initial and demographic data, show that increased incidence of TV in a population increases HIV incidence in the population. It is further shown that control strategies, such as treatment, condom-use and counselling of individuals with TV symptoms, can lead to the effective control or elimination of HIV in the population if their effectiveness level is high enough. / Dissertation (MSc)--University of Pretoria, 2017. / DST-NRF SARChI Chair in Mathematical Models and Methods in Biosciences and Bioengineering (M3B2) / Mathematics and Applied Mathematics / MSc / Unrestricted

Friction force oscillations caused by changing properties of the contact zone between brake disc and pad are well known from various applications. Resulting effects like brake judder are known phenomena in brake technologies and in the scope of various scientific work. A new measure to potentially reduce brake torque oscillations is the active compensation with the use of the control system of a self-energizing hydraulic brake (SEHB). New in comparison to traditional disc brakes is the fact that the brake torque is measured by the pressure in an additional supporting cylinder. Thus, the brake system is able to work in brake torque control mode. Within this paper a dynamic simulation model of the SEHB is shown and evaluated with measurement data achieved from a full scale test rig for railway applications. Based on the simulation model a pressure control strategy is developed to minimize brake torque oscillations of lower frequencies. The control parameters of the simulation are transferred to the experimental setup. Finally, simulation and experimental results are compared. Future work will deal with the development of control strategies to additionally minimize brake torque oscillations of the higher dynamics.

Load control strategies for wind turbines are used to reduce the structural wear of the turbine without reducing energy yield. Until now, these control strategies are almost exclusively built up-on linear approaches like PID and model-based controllers due to their robustness. However, advances in turbine size and capabilities create a need for more complex control strategies that can effectively address design challenges in modern turbines. This work presents WINDL, a load control policy based on a neural network, which is trained through model-free Reinforcement Learning (RL) on a simulated wind turbine. While RL has achieved great success in the past on games and simple simulation benchmarks, applications to more complex control problems are starting to emerge just recently. We show that through the usage of regularization techniques and signal transformations, such an application to the field of wind turbine load control is possible. Using a smoothness regularizer, we incentivize the highly non-linear neural network policy to output control actions that are safe to apply to a wind turbine. The Coleman transformation, a common tool for the design of traditional PID-based load control strategies, is used to project signals into a stationary coordinate space, increasing robustness and final policy performance. Trained to control a large offshore turbine in a model-free fashion, WINDL finds a control policy that outperforms a state-of-the-art controller based on the IPC strategy with respect to the prima-ry optimization goal blade loads. Using the DEL metric, we measure 54.1% lower blade loads in the steady wind and 13.45% lower blade loads in the turbulent wind scenario. While such levels of blade reduction come with slightly worse performance on secondary optimi-zation goals like pitch wear and power production, we demonstrate the ability to control the trade-off between different optimization goals on the example of pitch versus blade loads. To comple-ment our findings, we perform a qualitative analysis of the policy behavior and learning process. We believe our work to be the first application of RL to wind turbine load control that exceeds baseline performance in the primary optimization metric, opening up the possibility of including specialized load controllers for targeting critical design driving scenarios of modern large wind turbines.:Problem Method Aim Results Conclusion