Search

This thesis examines the physical legacy left by the unification of the Kingdom of England during the tenth century, and seeks to redress the way in which the Kingdom of Mercia is often overlooked or discounted in the traditional historical narrative. It principally examines the means by which Æthelflæd of Mercia extended political and military control over the West Midlands, both in terms of physical infrastructure and through ‘soft’ power in terms of economic control and material culture. It uses landscape archaeology, artefactual and textual evidence to compare Mercia with its ally, Wessex, and assess the different means by which Æthelflæd of Mercia and her brother Edward the Elder were able to consolidate and expand their territory, the physical infrastructure they established in order to defend it, and the ways in which these sites developed in response to the changing political, military and economic climates of the later tenth century. It will assess why some defensive sites developed into proto-urban settlements while others disappeared, and the extent to which this was a conscious or planned process. This thesis seeks to overturn the idea that burhs constructed in Mercia were insignificant or unplanned ‘emergency’ sites and instead were part of a sophisticated network of landscapes of defence, reflecting a significant level of manpower and logistical investment on the part of the Mercian state. It will furthermore seek to explore the ways in which the Mercian state supported such a network, how sites were chosen, constructed, maintained and garrisoned, and the impact these sites had both on the local population, in terms of patterns of settlement and material culture, and on the wider political scale.

An extensive study has been made of the factors affecting sample dispersion in flow injection analysis. This has included the effect of different manifold parameters on peak height and shape, and the use of various flow models to describe the physical processes involved. The factors investigated included sample flow-rate, tube length and diameter, sample volume, and the diffusion coefficient of the solute. The role of the detector has also been investigated, and was found to exert a strong influence on the observed dispersion behaviour of a flow injection manifold. In the light of this, a low volume flow cell was constructed which minimised the contribution of the detector to dispersion. Different modes of sample injection ("time" and "slug" injection) have been studied, and an automatic valve constructed in connection with the work. Both new and existing flow models have been evaluated in terms of their ease-of-use and ability to predict peak shape. Conditions for which the different models were valid were derived both experimentally and from consideration of the assumptions of the models. The applications of such models to the characterisation and design of flow manifolds has also been considered, and this approach compared to the purely theoretical and empirical approaches that have been described in the literature. The models examined in this thesis are the convection, diffusion, tanks-in-series, well-stirred tank, two-tanks in- parallel and two-tanks-in-series models. Finally, the practical implications of this work for the design of flow injection manifolds and the derivation of so-called "rules for dispersion" have been discussed, and suggestions made for further areas of research.

The liability of a steel to failure by brittle fracture may be judged by several criteria as exhibited in the various standard engineering tests. The relative influence of the many factors governing the ductile—brittle transition may vary greatly from test to test with the result that the various arbitrarily defined transition temperatures will also vary over quite a large range. Despite the complexity of the relationship between tests, some form of correlation is frequently demanded. This thesis presents experimental results in which, starting with the slow bend test on a standard V—notch Charpy specimen, a test has been successively modified, care being taken to alter only one variable at a time. In this way it was possible to study the relative effect of strain rate, size and method of stress application (i.e. notch—tension or notch—bending) on the change in transition temperature between tests. Load—deflection diagrams were obtained on all types of specimen so that the loads, energies, etc., at any given stage of the test, might be determined. Special instrumentation was developed to obtain diagrams of load—deflection, as well as of load—time, in the impact test. A comparison of these two diagrams allowed the variation of striker velocity in the impact test to be studied, and it concluded that the generally accepted assumption of the velocity varying linearly with time was likely to be in error, and that some form of shock loss of velocity occurred at impact. Finally the experimental results were examined in the light of some of the more recent hypotheses which have been advanced on the brittle fracture phenomenon.

The power system economic load dispatching task is formulated as a minimal-cost capacitated network flow problem. Then D.R. Fulkerson's 'Out-of-Kilter' method is used to solve the power system network flow model for the economic schedule of generation. Multiple-arc branch models form piecewise linear approximations to the quadratic cost of transporting power. Generator cost of production functions are easily and accurately modelled by one network arc per range of constant incremental cost. In a series of computational experiments on test systems of various sizes the validity, speed, and accuracy of the method was confirmed. Documentation of the FORTRAN computer programs written for the experiments is given. These are: the network flow solution algorithm, a conversational mode load flow program, and an economic scheduling program. A large power system with 177 generators was solved in 3 minutes of computing time on the IBM7094. Test data and results for smaller systems are included in the thesis.

Spring viraemia of carp virus (SVCV) is an OIE listed, notifiable disease affecting carp aquaculture. Disease outbreaks can lead to mass mortality events which poses significant economic threats to farmers and fishery owners. However, SVCV research has been limited due to the lack of full-genome sequences available for isolates from all genogroups, as well as limited information on SVCV pathogenicity. This study performed full-genome sequencing on several SVCV isolates from each of the recognised genogroups. Phylogenetic analysis of these sequences confirmed that isolates fall into 4 genogroups Ia-Id, which had been previously reported in literature. Interestingly, there was an additional sequence complementary to the grass carp thymosin beta-a gene in the 3' non-coding region of the G gene of a few of the Asian strains (Ia) that could influence the virus replication in that host. A conventional RT-PCR assay was designed to study the prevalence and distribution of the unusual additional sequence. Sequence alignments were used to design a new SVCV-specific TaqMan RT-qPCR assay based on the polymerase (L) gene. Testing of this assay on representative isolates from genogroups Ia-Id showed it was highly specific; detecting all SVCV isolates tested at low virus loads and did not cross-react with other closely related Spriviviruses. An in vitro investigation into differences in the innate immune response to isolates from different genogroups revealed that the Ia genogroup may be able to evade the immune response to promote their own replication whereas isolates from the Ib genogroup induced a strong anti-viral response that could play a role in reducing their pathogenicity. This study has also made good progress towards the development of a reverse genetics reporter system for SVCV and VHSV, incorporating green fluorescent protein (GFP) as a reporter gene. This information can be used going forward to develop a complete reverse genetics system for SVCV in which viable virus is generated from DNA, and then through genetic mutation allow us to learn more about the mechanisms involved in SVCV pathogenicity.

Electrification of the transport sector is a necessary measure required to lower greenhouse gas emissions. However, electrified light rail systems suffer energy efficiency problems due to the ineffective recovery of the braking energy regenerated during train or tramcar braking events. Utilizing the growing number of electric vehicles (EVs), and therefore the increasing number of EV batteries, as energy storage for the rail could provide a novel solution to address the challenge. Using tramcar operational data collected with a GPS data acquisition system, and the Matlab and Simulink modelling and simulation environment, this research initially gains insight into the energy balance of a local tram system, highlighting that the unrecovered and thereby wasted braking energy could be equivalent to approx. 50% of the overall energy consumption. This research later simulates the addition of a stationary energy storage system (SESS) to the tram network, and demonstrates the energy-saving achieved. Additionally, the simulation also suggests that a small but optimal capacity, and an installation on the mid-stop of an energy supply section, are beneficial to achieving better energy-saving and economic returns. Subsequently, this study designs two energy storage systems (ESSs), the EV energy storage system (EVESS), which solely exploits EV batteries for energy storage, and the combined ESS (CESS), which integrates the EVs with a sub-system of a stationary battery. Both ESS arrangements were found to successfully deliver energy-saving to the tram system. Moreover, a control approach was designed for the CESS to prioritise the charging of the EVs. The CESS could therefore refresh the capacity of the stationary battery more frequently and achieve a better overall energy-saving. Finally, this work compares the energetic and economic performance of the SESS and CESS addition to the network, and concludes that the CESS is more advantageous than the SESS alone. A network-wide CESS installation was therefore examined to assess its economic feasibility, and it was found economically viable if the electricity delivered to the EVs was charged at a reasonable price.

Today, the number of installed domestic PV systems is increasing, and more efficient appliances are used in households. This trend is a good signal for the UK Government, which has set a target to cut the UK CO2 emissions by 80% from 1990 levels before 2050. Further legislation was passed in 2019 to amend the target to 100% by 2050, to reach net zero emissions. However, with the increasing PV penetration, the level and direction of power flow on the UK electricity grid is less predictable, which brings challenges to current power grid and energy suppliers. Smart grid technology can solve the drawbacks of the PV penetration, and is the subject of significant investigation. In this thesis, hourly load profile modelling is introduced as the basis of the research, then a PV generation profile is determined for each typical size of PV system installed in the UK. An evaluation of the combined load and PV profiles throughout a year is carried out to address the sizing of an additional battery energy storage system. This facilitates an integrated domestic energy storage facility with renewable energy source, in order to create a win-win situation for both the customers and the grid. By using PV as an alternative energy source to power the home appliances, the system can reduce the dependence of household on grid energy, and it can shave the peak grid demand by managing the loads and exporting PV overproduction back to grid. Hence, the system can cut the electricity bill for customers and make profit by selling electricity to grid. The electricity tariff is considered when calculating the annual profit available, and conducting the system payback period for performance analysis. EV integration to the household in the form of vehicle to grid (V2G) is then examined based on the models developed. The complete domestic energy system model, including photovoltaic (PV) panels, battery energy storage system (BESS) and electric vehicle (EV) is updated to evaluate the impact of the V2G on the payback periods for a consumer. With a series of control algorithms applied, along with possible electricity tariffs, minimum energy usage and relative payback period for each variation of PV and BESS sizes within a proposed system are calculated. A battery ageing model is then developed to consider the annual battery degradation cost. It is shown that an EV can be used as extended stationary energy storage, together with a household BESS, to enhance the overall system performance.

This thesis discusses a transformerless multilevel converter (MLC) applied to a domestic level renewable energy system consisting of PV panels and EV batteries in their 2nd life applications. MLCs enable the use of conventional switching devices due to reduced voltage stress. Being able to produce a multilevel output voltage waveform, MLCs require less filtering and therefore may produce better quality waveform when compared to a standard 2-level voltage source converter (VSC). In this study, various modulation techniques for MLCs are implemented and the performance of the converter analysed regarding regulations and standards. The system is designed to have two-stage power conversion, including a DC-DC boost converter for adjusting each stage battery voltage, and maximum power point operation of the PV panels in each module. This provides a stable input voltage for the DC-AC converter stage. The cascaded H-bridge converter (CHB) is selected for the DC-AC conversion due to its isolated DC source requirement. This topology enables the separation of the total DC link voltage into different modules, increasing the accessibility of EV batteries in their 2nd life application. The base system is designed to be coupled without a transformer to the single-phase UK utility grid. A systematic approach is adapted for examining the MLC system. The design procedure starts with system parameter definition and component selection. This is then validated using simulation analysis and hardware implementation to demonstrate the practicability of the system for the planned application. The control algorithm is implemented in a National Instruments (NI) CompactRIO FPGA that can transform graphical programming into VHDL code. To accelerate the implementation and optimisation process, a co-simulation environment is used between NI LabVIEW and NI Multisim software. This ensures the optimisation of control code before compilation and enables testing without having analogue circuitry. Converters without galvanic isolation may exhibit ground leakage currents when coupled with grounded PV panels. This thesis analyses the common-mode and differential-mode voltages that CHB modules generate, and their effect on ground leakage current. The mathematical analysis suggests that leakage current may be supressed solely on changing the modulation method in a CHB converter. A novel leakage reduction pulse width modulation (LRPWM) technique is proposed, which successfully diminishes the ground leakage current to within the limit allowed by VDE-0126-1-1 (withdrawn, accessed in 2018) or IEC 62109-2 standard. The experimental results show that LRPWM has superior performance when compared to conventional MLC modulation techniques.

Frequency regulation is an important part of grid ancillary services in the UK power system to mitigate the impacts of variable energy resources and uncertainty of load on system frequency. The National Grid Electricity Transmission (NGET), the primary electricity transmission network operator in the UK, is introduced various frequency response services such as firm frequency response (FFR) and the new fast enhanced frequency response (EFR), which are designed to provide real-time response to deviations in the grid frequency. Flexible and fast response capabilities of battery energy storage systems (BESSs) make them an ideal choice to provide grid frequency regulation. This thesis presents control algorithms for a BESS to deliver a charge/discharge power output in response to deviations in the grid frequency with respect to the requisite service specifications, while managing the state-of-charge (SOC) of the BESS to optimize the availability of the system. Furthermore, this thesis investigates using the BESS in order to maximize triad avoidance benefit revenues while layering UK grid frequency response services. Using historical UK electricity prices, a balancing service scheduling approach is introduced to maximize energy arbitrage revenue by layering different types of grid balancing services, including EFR and FFR, throughout the day. Simulation results demonstrate that the proposed algorithm delivers both dynamic and non-dynamic FFR and also EFR to NGET required service specifications while generating arbitrage revenue as well as service availability payments in the balancing market. In this thesis, a new fast cycle counting method (CCM) considering the effect of current rate (C-rate), SOC and depth-of-discharge (DOD) on battery lifetime for grid-tied BESS is presented. The methodology provides an approximation for the number of battery charge-discharge cycles based on historical microcyling SOC data typical of BESS frequency regulation operation. The EFR and FFR algorithms are used for analysis. The obtained historical SOC data from the analysis are then considered as an input for evaluating the proposed CCM. Utilizing the Miner Rule's degradation analysis method, lifetime analysis based on battery cycling is also provided for a lithium-titanate (LTO) and lithium-nickel-manganese-cobalt-oxide (NMC) battery. The work in this thesis is supported by experimental results from the 2MW/1MWh Willenhall Energy Storage System (WESS) to validate the models and assess the accuracy of the simulation results.

This thesis is focused on the modelling and analysis of DC-DC converter topologies used for bidirectional charging of electric vehicles. Bidirectional converters are used in vehicle-to-grid (V2G) systems to allow bidirectional power transfer between the vehicle and the grid. Following the investigation in the literature review of potential converter topologies used in V2G applications and modelling techniques, this research proposes the application of the cyclic-averaging method for analysis of the Dual Active Bridge, 4th order resonant CLLC converter, and series compensated Inductive Power Transfer (IPT) converter. First, the cyclic-averaging method is applied for analysis of a phase-shift modulated Dual Active Bridge converter (DAB). For implementation of the cyclic analysis, the operation of the converter is first analysed using a Spice simulation to determine the system’s operation modes and duty cycles. The cyclic-averaging model is validated against a Spice simulation and employed to predict the converter’s output and to perform harmonic analysis of the inductor current. Following the analysis of the DAB, a 4th order CLLC converter is evaluated considering frequency and phase-shift modulations. The cyclic-averaging model is derived to model the behaviour of the converter’s output and state-variables in steady state. Additionally, a Fundamental Mode Approximation (FMA) model and a novel piecewise-linear state-variable model are also implemented for comparison. The models obtained are validated using Spice and, for the phase-shift modulated converter, experimental results. Finally, the series compensated IPT converter is analysed considering operation under phase-shift modulation. A FMA model is derived and, using circuit transformation, the state-variable and cyclic-averaging models previously defined for the CLLC converter are adapted for the analysis of the IPT converter. A prototype is built for validation of the cyclic model. Overall, for all converters analysed in this research, the cyclic-averaging method showed good performance with considerably fast execution and accuracy similar to Spice simulations.

Batteries are increasingly important in modern technologies. This is particularly true in the automotive sector, with hybrid vehicles using batteries to augment the traction power traditionally provided by the internal combustion engine. In such applications, one of the most important factors is the Dynamic Charge Acceptance (DCA) performance of the battery. This study investigates the standard method for establishing DCA performance and determines how the individual parameters of the test procedure and external factors influence the performance of lead-acid cells. This work identifies shortcomings of the standard test, which result in the true DCA performance being better than the standard test suggests. A series of modifications are proposed, which are shown to produce a more representative result. An investigation is performed to determine the effect of cell degradation on charge acceptance. This shows that the DCA test itself is not well suited to assessing the effects of degradation on DCA, and causes the results to appear worse than reality. The work also demonstrates that the usual methods of characterising degradation do not correlate well with DCA performance, and there is very little reduction in charge acceptance over the operational life of the cell. Investigations are undertaken into methods by which DCA performance may be improved. This shows that the application of ac ripple currents to batteries causes a significant increase in charge acceptance, and demonstrates how the frequency of the ripple is important in achieving the best results. This study also shows that the ripple currents have no detrimental effects on the health of the battery. Finally, the work is extended to cover lithium cells. This shows that whilst the DCA performance of lithium is more consistent, maximum charge acceptance is less than lead. It is shown that, by reducing maximum charge voltage, cycle life of cells can be extended without significant loss of stored energy.

This thesis describes novel techniques and developments in the design and implementation of a low power radio frequency (40kHz to 1MHz) wireless power transfer (WPT) system, with an application in the wireless charging of autonomous drones without physical connection to its on-board Battery Management System (BMS). The WPT system is developed around a matrix converter exploiting the benefits such as a small footprint (DC-link free), high efficiency and high power density. The overall WPT system topology discussed in this thesis is based on the current state-of-the-art found in literature, but enhancements are made through novel methods to further improve the converter's stability, reduce control complexity and improve the wireless power efficiency. In this work, each part of the system is analysed and novel techniques are proposed to achieve improvements. The WPT system design methodology presented in this thesis commences with the use of a conventional full-bridge converter. For cost-efficiency and to improve the converters stability, a novel gate drive circuit is presented which provides self-generated negative bias such that a bipolar MOSFET drive can be driven without an additional voltage source or magnetic component. The switching control sequences for both a full-bridge and single phase to single phase matrix converter are analysed which show that the switching of a matrix converter can be considered to be the same as a full-bridge converter under certain conditions. A middleware is then presented that reduces the complexity of the control required for a matrix converter and enables control by a conventional full-bridge controller (i.e. linear controller or microcontroller). A novel technique that can maximise and maintain in real-time the WPT efficiency is presented using a maximum efficiency point tracking approach. A detailed study of potential issues that may affect the implementation of this novel approach are presented and new solutions are proposed. A novel wireless pseudo-synchronous sampling method is presented and implemented on a prototype system to realise the maximum efficiency point tracking approach. Finally, a new hybrid wireless phase-locked loop is presented and implemented to minimise the bandwidth requirements of the maximum efficiency point tracking approach. The performance and methods for implementation of the novel concepts introduced in this thesis are demonstrated through a number of prototypes that were built. These include a matrix converter and two full WPT systems with operating frequencies ranging from sub-megahertz to megahertz level. Moreover, the final prototype is applied to the charging of a quadcopter battery pack to successfully charge the pack wirelessly whilst actively balancing the cells. Hence, fast battery charging and cell balancing, which conventionally requires battery removal, can be achieved without re-balance the weight of the UAV.

The combustion of fossil fuels, a non-renewable and finite resource, has caused increasing air pollution, ozone damage, acid rain and global warming. Electric vehicles (EVs) are eco-friendly with the attractive properties of lower greenhouse emissions, lower fuel usage and reduced air pollution. Battery and charger is one of the vital elements to analyse and develop for EVs. This research focuses on the bidirectional on-board battery charger with the emphasis on the design and analysis of the DC-DC converter. Firstly, a LLC resonant topology is selected as the initial design candidate of the DC-DC resonant converter owing to the preferred soft-switching features. Single phase chargers suffer from a second harmonic voltage/current ripple which can lead to reduction of battery life. A feedforward-proportional-integral-resonant (FF-PIR) controller has been proposed and tested for suppressing this low-frequency current ripple in the LLC resonant converter employed in EV battery chargers. Secondly, the recent developments in smart-grid technology necessitates bi-directional power flow from distributed energy storages like EV to support grid in the vehicle-to-grid (V2G) application. To achieve bi-directional power flow capability, the conventional uni-directional LLC topology is modified into bidirectional CLLC resonant converter. The characteristics have been analysed and validated by extensive simulations and experimental tests. Further improvement has been implemented to increase the power efficiency under varying battery voltages. An optimum-resonant-frequency tracking scheme is proposed and tested to maintain the operation close to the maximum efficiency point over a wide battery voltage range. Finally, in order to predict the converter efficiency accurately, this thesis presents a prediction method employing 2D and 3D finite element analysis (FEA) for calculating the power losses of magnetic components with litz wire in the converter.

A variety of economic factors currently motivate the development of electrochemical energy storage. The effective use of renewable energy requires short term storage, for which electrochemical cells may be used. Electrified transport is also driving development; stored energy limits the range of electric vehicles. In hybrid vehicles, improved dynamic charge acceptance will help to optimise powertrain efficiency. A non-invasive measurement of current distribution within a cell is a useful aid to understanding its operation and optimising its design. Here, the coupling between the cell current and the resulting magnetic field is exploited by taking measurements of magnetic flux density outside the cell and inferring the current distribution within. This technique may be termed magnetic tomography or magnetotomography. In this thesis, a practical system is implemented in order to observe the current distribution within a single lead acid cell. An existing method of constraining and solving the inverse problem is adapted for use in conjunction with 3D finite element software, to make it suitable for modelling the complex geometry of a commercial electrode. Some tolerance of unknown material conductance is built into the solver method. An array of sensors is used to obtain a set of magnetic field measurements simultaneously, allowing temporally- and spatially- resolved current distribution images. Solutions from the magnetic tomography system are verified against data from an array of ferrous cores, submerged in the electrolyte. Measurements are taken while the cell is operated at a current of approximately 0.625 C. The current distribution is found to be very uniform throughout most of the testing, although fatigue of the cell plates does lead to a non- uniform distribution. The magnetic tomography system is tested on both uniform and non- uniform distributions. Mean absolute errors of approximately 5 – 7 % are achieved. The effect of model errors on solution accuracy is investigated.

This thesis describes multilevel converters (MLC) designed for use with a repurposed electronic vehicle battery (battery 2nd life). MLC is of particular interest due to the low harmonic distortion content and reduced voltage stress in the switching devices. A detailed study of the MLC topologies and modulation techniques is presented. Space vector modulation is analysed and implemented to evaluate the converter. A comprehensive assessment of the MLC is presented using wide bandgap (WBG) devices highlighting the devices’ thermal and high switching frequency features.

This thesis sets out a series of new techniques to improve the thermal management of power electronics. The work is motivated by the increasing impetus to design smaller, more energy efficient electronic power systems for a range of applications, notably electric vehicles. Thermal management is an increasingly important tool which can facilitate improvements in power density through better monitoring and control of system temperatures. This thesis seeks to deliver improvements in implementing this strategy. A review of the state of the art in thermal management is reported, focussing on temperature measurement, thermal characterisation and system modelling techniques. In addition, novel techniques for arbitrary dissipation control and die temperature measurements in semiconductor devices are presented. A novel analysis of the limitations of low-order thermal models is also described. Improvements and applications of these techniques form the basis of this thesis. The pseudorandom binary sequence (PRBS) technique for system identification is applied throughout the thesis to characterise thermal systems. A mathematical analysis is provided, together with a novel technique to determine the minimum gain which can be identified by PRBS techniques in the presence of noise. A novel improvement to the PRBS technique for typically ten times more noise resilient measurements is then developed based on mathematical mixing of different frequency PRBS signals. In parallel, a novel technique is formulated to estimate the temperature throughout a multiple device system using digital IIR filters and PRBS thermal characterisation, which achieves errors of 3-5% when demonstrated practically. By combining these techniques, a comprehensive temperature estimation and control methodology is implemented for a multiple device system under active cooling. Finally, the expansion of the proposed methodologies to steady-state die temperature estimation is presented with comparable accuracy to surface temperature measurements, increasing the usefulness of the developed techniques in a practical setting.

The research objective was to investigate the generation and collection of power generated in a distributed tidal generation site, and realise the most economic grid connection methodology for a practical installation. Current tidal steam prototype designs utilise variable speed gear boxes, power frequency converters and variable pitch blades in the nacelle, thus requiring complicated, heavy and expensive equipment in the nacelle. Moreover the variable blade pitch mechanism has a large power demand, and has to operate day after day submerged in water at depths between 30m – 80m. The approach taken in this research is to replace major cost components with much cheaper ones, as well as locating as many components as possible onshore where costs are less. This thesis investigates a novel tidal stream power generation system referred to as “passive rectification to a common DC-bus” where an array of 3-phase synchronous generators operating at diverse speeds are connected to a common DC bus via passive diode rectifiers. The proposed tidal stream topology shows that variable pitched blades, variable speed gearboxes and frequency converters in the nacelle can be replaced by a much simpler system that uses fixed pitch blades, a fixed ratio gearbox, generator excitation control system, and a diode bridge rectifier. These components are generally cheaper, lighter, have a lower power demand and are more reliable than the components they replace. Also some of the voltage and frequency conversion is carried out ashore where costs are less, in addition to the overall reduction in the number of components required as some of the components are shared by multiple devices. As a result system efficiency is maximised, in addition to facilitating a reduction in capital, installation and operational costs. Power conversion from the tidal generation power system to shore is achieved via passive rectification to a common DC-bus, thus system power regulation is achieved by individual generator field current control, to attain optimum system operation including, maximum power extraction, power limitation and stall control, in turn improving system reliability and controllability. Furthermore passive stall control via generator field current regulation eliminates the need for mechanical brakes which reduces cost and improves reliability. The feasibility of the proposed power generation system was carried out via computer simulations and later validated via small scale laboratory hardware simulations.

This thesis details novel control schemes and design techniques with the aim of improving the performance of several switched-mode power converter topologies. These improvements include higher steady-state and transient efficiencies for hard-switching converters and the automatic current limiting provision for LLC resonant converters. The thesis initially attempts to use linear closed-loop controllers to improve the transient response of synchronous buck converters, enabling them to be designed with a lower open-loop bandwidth so that the system can achieve higher efficiency. Three types of controllers were investigated viz: the PID, the state-feedback and the predictive controller. All three controllers exhibit similar step responses, which are the maximum transient responses achievable by the linear controllers with the given requirements. The thesis then examines the parallel converter (i.e. a converter with two parallel connected power modules (PMs)) in detail with a view to improve the efficiency and to minimise the current ripple experienced by the output capacitor. Two control schemes and a design technique for the parallel converter are proposed, to simultaneously improve its efficiency and power density. The parallel converter in this research consists of two non-identical rated PMs (termed main PM and auxiliary PM), with the transient response requirement allocated to the auxiliary PM, thereby allowing the main PM to operate at a lower frequency for higher steady-state efficiency. The first control scheme activates the auxiliary PM only when a pre-determined deviation in load/output voltage is exceeded under a load step. Thus, eliminating the losses contributed by the low efficiency auxiliary PM for small load step changes. The second control scheme shapes the auxiliary PM inductor current to be equal and opposite to the main PM current ripple, which when combined reduce the current ripple as experienced by the output filter capacitor, thereby allowing a lower value (and hence physically smaller) capacitor to be selected for higher power density. In order to improve the converter's steady-state efficiency further, the minimum load condition is allocated to the auxiliary PM in the new design technique. These allow both the main PM inductance and its switching frequency to be lower for higher efficiency. In recent years, the LLC has received much attention owing to its favourable operating characteristics including high efficiency and high power density. Usually one chooses to operate at or very close to the load independent point (LIP) since very little control effort is required to regulate the converter's output voltage in response to changes in the load. However under fault conditions where the load tends towards a short circuit, excessive currents can flow and thus control action need to be taken to protect both the converter and the load. The final topic of the thesis hence studies the characteristics of an LLC resonant converter with current-limiting capacitor-diode clamp and develops a new equivalent circuit model to predict the behaviour under overload conditions. A detailed analysis of the converter is presented using the proposed model, from which a design methodology is derived allowing the optimum circuit components to be selected to achieve the required current limiting/protection characteristics.

This thesis investigates how a back-to-back connected, 5-level, diode-clamped multilevel inverter may be applied in an aerospace Starter / Generator (S/G) system. The performance of the proposed system is investigated with both wound field (WF) and permanent magnet (PM) synchronous machines. Control techniques are investigated to improve the waveform quality at the machine side. Also the grid interfacing of the converter is considered to make the system more resilient to AC distribution bus distortion. Aerospace generators are typically low inductance machines owing to the need to regulate the back-emf at the high rated speed, reduce stator losses and to increase power density. However, the driving of low inductance synchronous machines with a two level inverter results in a high level of current ripple at the same frequency as the converter switching frequency. A high level of current ripple increases losses in both the converter and machine necessitating additional heatsinking/cooling. The high ripple also results in increased radiated and conducted EMI. Additional filtering/shielding is therefore required to protect other vulnerable systems in the vicinity. The stability of the converter's controller may be adversely affected by a high level of current ripple. This instability can lead to the increased production of low order harmonic in the current waveform. The effect of interfacing a two-level inverter to an aircraft's AC distribution bus must also be considered. Power quality standards require the converter to draw a high quality sinusoidal current from the grid and introduce minimal distortion onto the voltage waveform. In order to reduce the current ripple in the machine windings and to ensure the distortion at the grid connection is within the governed limit, it is customary to use increased passive filtering and/or raise the converter switching frequency. Both of which incur a weight penalty. This study sought to investigate how a multilevel converter may be used as an alternative approach to this problem. The stepped PWM waveform produced by a multilevel converter offers superior harmonic performance and may therefore require reduced amounts of passive filtering whilst still producing a high quality waveform. The performance of the converter is initially demonstrated experimentally upon a PMSM. This is considered relevant as future 'embedded' S/Gs are likely to be PMSM due to the robust rotor structure and high power density. Performance is verified experimentally using a dynamometer test rig. The dynamometer is controlled to provide a rudimentary model of a gas turbine in both starting and generating mode. It is therefore possible to load the S/G in a similar manner to a real world application and provide realistic testing data. The converter is then applied to a WFSM. WFSMs are the industry norm for aerospace generators and it is this machine topology that is used as the S/G in Trent 1000 engine. Converter performance with a WFSM must therefore be assessed if this study is to be considered relevant. The excitation of the machine is investigated to ensure that the machine may be driven in motoring mode from zero speed and to maximise the available torque throughout the starting sequence. The excitation is then used to regulate the back-emf during generating mode. The same dynamometer test rig is used to provide full system test experimental data. Predictive control is then investigated to obtain a further reduction in waveform distortion in the stator current. Predictive control utilises a mathematical model of the load to predict the converter output voltage required in the next PWM period to minimise the current error. The estimated switching vector may then be applied in the next modulation period. This is shown to provide a substantial reduction in THD throughout both modes of operation. A reduction in THD in a real-world application would provide lower losses in both the machine and converter. The heatsinking/cooling arrangement may be reduced giving a possible weight reduction. In the final stage of this study, the control and synchronisation of the grid connected inverter is investigated. Aerospace power converters are required by the standards to function under distorted grid conditions. In this study a review is conducted into the competing grid synchronisation methods, to determine which is the most suitable to accommodate the heavily unbalanced grid waveforms commonly found on an aircraft distribution bus. A virtual-flux estimator is identified as the most promising technique as it eliminates the three voltage transducers at the grid connection. This not only makes the converter more resilient to grid distortion but may also increase the converter's reliability as three essential components have been removed. The system is assessed in both simulation and hardware. The level of grid distortion applied to the grid waveform is the maximum permissible according to the aerospace power quality standards. The system performance is satisfactory, maintaining synchronisation with the grid with minimal error despite the heavily unbalanced supply.