The radial force control has been widely investigated for a variety of machine topologies aiming at the reduction of the mechanical stress on bearings as well as the overall vibration. Furthermore, it can be also applied to bearingless drives where the force control can be exploited to suspend the rotor element. Traditional radial force control techniques rely on two independent sets of windings dedicated to torque and suspension force, respectively. However, different winding structures can be employed to reduce the overall system complexity. Multiphase machines, in particular, offer the possibility to embed the radial force and torque generation in a single winding set and possess an enhanced fault tolerant capability. This thesis presents an alternative radial force and torque control technique for a multiphase sectored permanent magnet synchronous machine. The mathematical model of the machine and the theoretical investigation of the force production principle are provided. A force control methodology based on the minimization of the copper losses is described and adopted to calculate the d-q axis current references. The predicted performances of the test machine are benchmarked against finite elements analysis. An experimental rig has been manufactured in order to validate the radial force and torque control. The experimental validation focuses on the suppression of selected vibration frequencies at different speeds. The radial force control has been also exploited to levitate the rotor in order to obtain a two degrees of freedom bearingless drive. The fault tolerant capabilities have been investigated and experimentally tested considering the open-circuit of one three-phase winding. The tuning of the x-y axis position controller has been investigated and a state space approach has been followed to synthesize different position controllers.

Solid rotor induction machines are suitable for applications which require robustness, reliability and high rotational speed. A literature review of high speed technologies is initially presented. The current limitation and challenges are detailed based on a wide collection of data. The multi-physics aspect related with electrical machines for high speed applications are discussed providing a summary of the current state of the art. The main aim of the research was to develop a multi-physic computational environment for the design and analysis of solid rotor induction machines. The electromagnetic, thermal, structural and rotor dynamics models have been developed targeting reduced computational time and accurate predictions. Numerical techniques are proposed based on the discretisation of the computational domain. The different disciplines are linked together providing a flexible and powerful tool for the characterisation of solid rotor induction machine. Another objective was to investigate the impact of the rotor material on the electromagnetic performances of the machine. Finite Element simulation are used to account for the non linear magnetic properties. The impact on the equivalent circuit parameter is discussed and general criteria for material selection presented. Three dimensional finite element calculation are p erformed targeting the validation of the end region correction factor and select the rotor length. The performances of a 120 [kW]−25000 [rpm] solid rotor machine are compared with a caged rotor induction machine for waste heat recovery application.

This study explored the effects of mechanical loads on Li-ion pouch cells by considering their operation under laboratory conditions. The focus of the research conducted was exploratory in nature with the aim of developing advanced experimental methods and techniques to answer a specific research question motivating this work: \Do mechanical events influence the electrochemical performance of Li-ion batteries?". In order to address this research question, the following goals were targeted: 1) study the effects of high g impacts on the cell performance and investigate the extent of cell damage under such an event, 2) examine the influence of mechanical bending loads on cells and ageing effects introduced on the cell performance. In the context of studying the effects of high g pulses and mechanical bending load on the electrochemical performance of Li-ion batteries, a comprehensive analysis of the internal impedance and capacity measurements was undertaken. Throughout the entire study, none of the analyses established any signs of statistically significant relationships between the specified electrochemical parameters. This study therefore finds that high g pulses and external mechanical bending load have no adverse influences on the electrochemical characteristics of Li-ion batteries in use, within the bounds of the investigation, as no evidence of electrochemical performance degradation by the effects of such events were substantiated. The study examined the influence of charge/discharge cycles on the load relaxation characteristics of a cell retained under bending deformation, by quantifying its structural evolution prior to and post electrochemical cycling using X-ray CT. It was ascertained that a cell subjected to a constant bending deformation during electrochemical cycling experienced a healing effect, owing to its viscoelastic properties and volume expansion of the electrodes during charging and discharging. The work shows tantalising evidence that external mechanical load on a cell may provide possibilities to improve its electrochemical characteristics. It is recommended that this phenomena is investigated further.

In recent years, research in the area of multiphase drives has increased significantly. Having higher number of machine phases allows the current to be shared between the phases, thus reducing the current rating of power semiconductors used in the power converter. Additionally, if a multilevel inverter is used to drive the machine, the output voltage waveforms are going to be approximated closer toward sinusoidal waveforms, thus resulting in lower total harmonic distortion. Therefore, the combination of multiphase and multilevel technologies gives considerable benefits compared to conventional two-level three-phase drives. Unlike a carrier-based approach, which can be easily expanded to any number of converter voltage levels and any number of machine phases, the development of space vector algorithms is also reliant on the machine’s configuration. In other words, different drive topologies require their own unique space vector algorithms. In fact, the complexity of developing a space vector algorithm will dramatically increase with the increase of number of levels and/or number of phases. This thesis presents pulse width modulation techniques for two- and three-level asymmetrical and symmetrical six-phase drives with a single or two isolated neutral points configuration. However, since the modulation techniques for the drives with two isolated neutral points are based on the well-established modulation techniques for three-phase drives, more emphasis is given towards the development of modulation techniques for single neutral point case, particularly those that are based on space vector algorithm principles. In order to realise sinusoidal output phase voltage waveforms, several requirements and conditions have to be met. The requirements revolve around ensuring that the low order harmonics, which contribute to the machine losses, will not exist. Meanwhile, the conditions are more towards minimising the switching losses. All modulation techniques are verified through simulation, while those for three-level case are validated experimentally as well. Comparison and discussion of obtained simulation and experimental results, performance and complexity in terms of execution time of the developed modulation techniques, are presented. The equivalence between corresponding modulation techniques, which are based on the space vector algorithm and carrier-based approach are also established.

Due to the rapid development of the power electronics in the second half of the twentieth century, a significant research effort has been put into the modelling of electrical machines to provide mathematical models for control purposes. As the power electronics isolate the machine from the grid, the number of phases on both sides no longer needs to be the same, thus allowing for use of multiphase machines. Several studies have shown that multiphase machines can yield lower torque ripple, provide higher torque per phase current, and that they can continue to operate with one or more faulty phases, thus increasing the robustness of the power stage. This, amongst other benefits, has led to increased interest in multiphase machine employment for critical applications, such as more-electric aircraft, electrical propulsion systems for ships and offshore wind, etc. Amongst the different multiphase machine constructions, the multiple three-phase winding structure with isolated neutral points is of special interest. It can be operated using multiple three-phase converters, so that almost no modification of hardware is needed. Furthermore, with high power machines (above the 5 MW class), several converters in parallel should be used when increased availability is desired. This is where multiple three-phase winding machines show an additional benefit, galvanic isolation between the windings. By connecting one three-phase converter to each of the three-phase windings of the machine, the increased availability of paralleling converters is obtained while the problem of the circulating current between paralleled converters is practically eliminated thanks to said galvanic isolation. The control schemes of three-phase machines should not be directly applied to multiple three-phase winding machines, since these show internal cross couplings between the different three-phase windings that may affect dynamic performance. To examine the behaviour and design control schemes for multiple three-phase winding machines, modelling approaches based on vector space decomposition, multiple dq modelling approach and a novel approach, specifically developed in this thesis for the independent power flow control in individual three-phase windings, are studied. It is demonstrated that, by including appropriate decoupling terms in the traditional three-phase control structure, a completely decoupled operation can be obtained in all the three-phase windings in the machine when control scheme is based on the multiple dq modelling approach. With this control approach, the control of these machines is accomplished using control structures and model transformations familiar to those skilled in the art of the three-phase machines. For six-phase machines the existing transformations are sufficient for all control purposes, while the novel transformation becomes a useful tool when there are three or more three-phase windings. The influence of a low switching to fundamental frequency ratio on behaviour of the controlled object is also covered in this work. This has a great impact on the modelling of current control loops, especially when using the synchronously rotating reference frame in variable fundamental frequency applications, such as motor drives. The precise modelling of the actual control loops is of vital importance since it allows development of faithful control tuning techniques. With these, the regulator parameters, which ensure certain specified dynamic performance of the loops, are obtained and their behaviour can be precisely described and predicted by simulations. The machine’s parameter identification has also been approached in this work; accurate parameter knowledge is of essential importance to ensure the correct match between experimental and simulation results. All the experimental work has been done using a 150 kW permanent magnet synchronous generator in six-phase configuration with two three-phase winding placed spatially in phase. Unequal power sharing between different three-phase windings is studied further, including the simultaneous operation of one winding in motoring and the other in generation for a six-phase machine. This particular mode of operation has been found as very useful in development of a novel testing method for the machines with multiple three-phase windings, of synthetic loading type, which is fully verified by experimentation. A corresponding theoretical/simulation work has been performed for a nine-phase (triple three-phase) machine.

Bias Temperature Instability (BTI) is one of the most important reliability concerns for Metal Oxide Semiconductor Field Effect Transistors (MOSFET), the basic unit in integrated circuits. As the development MOSFET manufacturing technology, circuit designers need to consider device reliability during design optimization. An accurate BTI lifetime prediction methodology becomes a prerequisite. Typical BTI lifetime standard is ten years, accelerated BTI tests under high stress voltages are mandatory. BTI modelling is needed to project BTI lifetime from high voltages (accelerated condition) to operating voltage. The existing two mainstream BTI models: 1). The Reaction-Diffusion (R-D) framework and 2). The Two-Stage model cannot provide accurate lifetime prediction. Quite a few fitting parameters and unjustifiable empirical equations are needed in the R-D framework to predict the lifetime, questioning its predicting capability. The Two-stage model cannot project device lifetime from high voltages to operating voltage. Moreover, the scaling down of MOSFET feature size brings new challenges to nano-scale device lifetime prediction: 1). Nano-scale devices’ current is fluctuating due to the impact of a single charge is increasing as MOSFET scaling down, repetitive tests need to be done to achieve meaningful averaged results; 2). Nano-scale devices have significant Device-to-Device variability, making the lifetime a distribution instead of a single value. In this work a comprehensive As-grown Generation (A-G) framework based on the A-G model and defect centric theory is proposed and successfully predicts the Time Dependent Variability and lifetime on nano-scale devices. The predicting capability is validated by the good agreement between the test data and predicted values. It is speculated that the good predicting capability is due to the correct understanding of different types of defects. In the A-G framework, Time Dependent Variability is experimentally separated into Within-Device Fluctuation and the averaged degradation. Within-Device Fluctuation can be directly measured and the averaged degradation can be modelled using the A-G model. The averaged degradation in the A-G model contains: Generated Defects, As-grown Traps and Energy Alternating Defects. These defects have different kinetics against stress time thus need separate modelling. Various patterns such as Stress-Discharge-Recharge, multi-Discharging-based Multiple Pulses are designed to experimentally separate these defects based on their different charging/discharging properties. Fast-Voltage Step Stress technique is developed to reduce the testing time by 90% for the A-G framework parameter extraction, making the framework practical for potential use in industry.

Demand for the use of online services such as free emails, social networks, and online polling is increasing at an exponential rate. Due to this, online service providers and retailers feel pressured to satisfy the multitude of end-user expectations. Meanwhile, automated computer robots (known as ‘bots’) are targeting online retailers and service providers by acting as human users and providing false information to abuse their service provisioning. CAPTCHA is a set of challenge/response protocols, which was introduced to protect online retailers and service providers from misuse and automated computer attacks. Text-based CAPTCHAs are the most popular form and are used by most online service providers to differentiate between human users and bots. However, the vast majority of text-based CAPTCHAs have been broken using Optical Character Recognition (OCR) techniques and thus, reinforces the need for developing a secure and robust CAPTCHA model. Security and usability are the two fundamental issues that pose a trade-off in the design of a CAPTCHA. If a CAPTCHA model were too difficult for human users to solve, it would affect its usability, but making it easy would risk its security. In this work, a novel CAPTCHA model called VICAP (Visual Integration CAPTCHA) is proposed which uses trans-saccadic memory to superimpose a set of fleeting images into a uniform image. Thus, this will be creating a meaningful picture of the object using the sophisticated human visual system. Since the proposed model is based on this unique ability of humans, it is logical to conclude that none of the current computer recognition programmes has the ability to recognise and decipher such a method. The proposed CAPTCHA model has been tested and evaluated in terms of usability and performance in laboratory conditions, and the preliminary results are encouraging. As a result of this PhD research, the proposed CAPTCHA model was tested in two scenarios. The first scenario considers the traditional setup of a computer attack, where a single frame of the CAPTCHA is captured and passed on to the OCR software for recognition. The second case, implemented through our CAPTCHA-Test Application (CTA), uses prior knowledge of the CAPTCHA design. Specifically, a number of frames are individually captured and superimposed (or integrated) to generate output images as a single image using the CTA and then fed into the OCR programme. The second scenario is biased because it also requires prior knowledge of the time interval (ISI) to be used in the integration process. When the time interval is set to a value higher than the optimal ISI, there is insufficient information to complete the CAPTCHA string. When the time interval for integration is set to a value lower than the optimal one, the CAPTCHA image is saturated due to the uniform nature of the noise process used for the background. In order to measure the level of usability of our proposed VICAP model, a user evaluation website was designed to allow users to participate in the proposed VICAP model. This evaluation website also enabled participants to compare our proposed VICAP model with one of the current popular Google CAPTCHA models called ReCAPTCHA. Thus, to ensure the usability of the proposed CAPTCHA model, we set the threshold for the ORO (Original to Random Output Data) parameter at 40%. This ensured that our CAPTCHA strings would be recognised by human observers at a rate of 100%. In turn, when examining the robustness of our VICAP model to computer programme attacks, we can observe that for the traditional case of OCR recognition, based on a single-frame scenario, the Computer Recognition Success Rate (CRSR) was about 0%, while in the case of a multi-frame scenario, the CRSR can increase to up to 50%. In the unlikely scenario of an advanced OCR software attack, comprising of frame integration over an optimal time interval (as described above), the robustness of the VICAP model for the multi-frame sequence reduces to 50%. However, we must stress that this latter scenario is unfairly biased because it is not supported by the capabilities of present state-of-the-art OCR software.

The study in this dissertation aimed to develop novel slot waveguide and resonator based compact integrated photonic sensors. Using guided photons as the probe for detection and measurement, and finally converting the signal magnitude from any domain to an electronic signal is one of the most effective approaches of sensing. Novel and efficient designs of hybrid and composite plasmonic horizontal slot waveguides and dielectric straight slot resonators are proposed and optimised to detect a small refractive index change. Practically, the concentrations of chemical liquid and gas or vapour are often expressed in terms of ‘grams per litre (g/l)’ and ‘parts per million (ppm)’, respectively. Thus, device sensitivities related to the detection of those substances may be expressed in amplitude or wavelength shift of the output optical signal per unit g/l or ppm. However, changes in the concentration and the chemical property of any liquid, gas, and chemical vapour result in refractive index variation of those substances. Therefore, we emphasised the detection of the refractive index change which, on the other hand, represents the concentration and/or chemical property change in the testing sample. Design, optimisations, and performance analyses of those waveguides are carried out by a direct divergence modified full-vectorial two-dimensional (2D) finite element method (FV-FEM). It provides an accurate spurious free better characterisation approach to handle all types of waveguides especially, the plasmonic and hybrid plasmonic waveguides where the guided mode is a complex mixture of the dielectric waveguide mode and surface plasmon polaritons (SPPs). Additionally, a full-vectorial three-dimensional (3D) FV-FEM dedicated to solve the 3D resonator problems is also developed and implemented. As an application of the 2D FV-FEM, first a metal nano-wire with identical and non-identical cladding conditions are considered and modal evolutions of its plasmonic fundamentals and complex supermodes are studied which also work as a benchmark of the direct divergence modified 2D FV-FEM code. Different mode effective area definitions are incorporated with this newly modified code and the low and high index contrast and hybrid plasmonic complex waveguides are simulated to determine the appropriateness of different effective area approaches for various waveguiding structures. Following this, the 2D FV-FEM is implemented in designing complex plasmonic slot based sensing waveguides. A horizontal slot composite plasmonic waveguide structure with a low index porous ZnO (P-ZnO) layer as slot material is reported and also incorporated in a compact symmetric Mach-Zehnder interferometer (MZI) to detect the presence of ethanol vapour in the environment. The waveguide is optimised to obtain a maximum slot confinement (41%) and overall a high phase sensitivity of the MZI device. A similar hybrid plasmonic horizontal slot waveguide is designed and optimised for detection of small refractive index change in the bio-layers (ssDNA and dsDNA) during DNA hybridisation. Next, a metal strip loaded horizontal slot hybrid plasmonic waveguide is designed for a high slot confinement and lower modal loss. The waveguide structure contains a suspended Si slab on top of an optimised thin metal layer (silver) to obtain a lower modal attenuation. It shows an enhanced 60% and 82% power confinement in the slot and sensing (slot+clad) sections, respectively with a small modal attenuation value of 0.036 dB/um. This waveguide is incorporated in an asymmetric Mach-Zehnder interferometer with an asymmetric power splitting scheme which results in an improved interferometric fringe visibility. This compact device exhibits a high temperature and chemical concentration sensitivity of 244 pm/±C and 437.5 nm/RIU, respectively. Beside these waveguides, a silicon-on-insulator (SOI) based vertically slotted straight resonator is also reported in this thesis. Due to its easy and straight structural design it is free from the bending losses and its fabrication steps are much easier compared to other complex devices such as ring, disk resonators, and grating based sensors. The slot cross-section is first optimised and then its length is calculated with those optimised parameters. The 3D straight resonator as a whole is then considered for bulk and surface sensing. Complete performance analyses and the resonating wavelength shift of the device due to small refractive index change during bulk and surface sensing applications are determined by using the newly developed 3D FV-FEM code. This straight resonator exhibits a 5.2 nm resonating wavelength shift for a 5 nm ultra-thin bio-layer and high bulk sensitivities of 820 nm/RIU and 683 nm/RIU for filled and empty slot conditions, respectively.

The electric potential sensor is a novel, ultra high impedance sensor, previously developed at the University of Sussex. These sensors have been applied to a range of applications, including electrophysiology, non destructive testing of composite materials and novel nuclear magnetic resonance NMR probes. Some of these measurements can be made in a strongly coupled (≥100pF) mode, where the coupling capacitance is reasonably large and well dened, and ambient noise is therefore less problematic. However for many applications, there exists a requirement for this coupling to be much weaker. This weak and poorly dened coupling creates substantial problems with ambient noise often causing sensors to saturate and become unusable. In the past, therefore, these measurements have all been made inside electrically screened rooms and enclosures. The work discussed in this thesis explores the possibility of operating these sensors outside of electrically screened environments. A number of techniques for resilience against noise are explored and experiments to fully analyse and characterise the performance of the sensors are discussed. As a result of this work, further results are then shown for a number of experiments carried out in a busy lab environment, in the presence of noise sources, and with little or minimal screening used. In this case, data is shown for the collection of remote cardiac and respiratory data, imaging of the spatial distribution of charge on insulating materials, detecting electric eld disturbances for movement sensing and early results for a microscopic XY scanning application.

The work discussed in this thesis explores the possibility of employing the Electric Potential Sensor for use in healthcare and assistive technology applications with the same and in some cases better degrees of accuracy than those of conventional technologies. The Electric Potential Sensor is a generic and versatile sensing technology capable of working in both contact and non-contact (remote) modes. New versions of the active sensor were developed for specific surface electrophysiological signal measurements. The requirements in terms of frequency range, electrode size and gain varied with the type of signal measured for each application. Real-time applications based on electrooculography, electroretinography and electromyography are discussed, as well as an application based on human movement. A three sensor electrooculography eye tracking system was developed which is of interest to eye controlled assistive technologies. The system described achieved an accuracy at least as good as conventional wet gel electrodes for both horizontal and vertical eye movements. Surface recording of the electroretinogram, used to monitor eye health and diagnose degenerative diseases of the retina, was achieved and correlated with both corneal fibre and wet gel surface electrodes. The main signal components of electromyography lie in a higher bandwidth and surface signals of the deltoid muscle were recorded over the course of rehabilitation of a subject with an injured arm. Surface electromyography signals of the bicep were also recorded and correlated with the joint dynamics of the elbow. A related non-contact application of interest to assistive technologies was also developed. Hand movement within a defined area was mapped and used to control a mouse cursor and a predictive text interface.