Successful deposition of ZnSe and ZnS„Sei_x layers has been performed with close-spaced thermal evaporation (CSTE). ZnSe (Eg = 2.67eV) and ZnS,Sei, (Eg to 3.6eV) have the potential to replace CdS (Eg = 2.42 eV) as a buffer layer in solar cell applications, giving a two-fold benefit: (i) increased blue response of the cell, potentially allowing more light to reach the pn junction and contribute to photogeneration and (ii) reduce the toxic Cd element of the buffer layer. CSTE has produced films in which the deposition parameters can be controlled to alter the morphology of the as-deposited coatings. SEM and AFM investigations have shown that pinhole free ZnSe and ZnS„Sei_x films can be produced with this deposition process. In addition, the ZnSxSei, layers show a linear shift in lattice constant and a systematic shift in energy bandgap with alloy composition. XRD data and the steep absorption edges in the transmittance data confirm the good crystallinity of the layers. To partner the ZnSe and ZnS,,Sei_x buffer layers in a thin film heterojunction, CdTe absorber layers were grown in the superstrate configuration. These CdTe layers were deposited in the same deposition chamber, without breaking vacuum, to reduce the risk of interfacial contamination. ZnSe and ZnSxSei_x / CdTe solar cells were fabricated with the best cell producing PV characteristics of: short circuit current 17mAcm-2, open circuit voltage 460mV and efficiency approaching 3%. The spectral response of all ZnSe and ZnSxSei_x / CdTe devices demonstrated a systematic shift to shorter wavelengths with increasing alloy composition, therefore showing the potential of these materials to increase solar cell efficiency. This low cost deposition process has shown excellent potential to be scaled up for commercial applications.

An electrocardiogram (ECG) is a bioelectrical signal which records the heart‘s electrical activity versus time. It is an important diagnostic tool for assessing heart functions. The interpretation of ECG signal is an application of pattern recognition. The techniques used in this pattern recognition comprise: signal pre-processing, QRS detection, creation of variables and signal classification. In this research, signal processing and programs implementation are based in Matlab environment. The processed simulated signal source came from the SIMULAIDS® interactive ECG simulator™ device and the actual heart signals came from actual patients that suffer from various heart disorders, as well as healthy persons that hadn‘t recorded any form of heart condition in the past. For the creation of the database in this research, 5 types of ECG waveform were selected from the ECG simulator device. These are normal sinus rhythm (NSR), ventricular tachycardia (VT poly), ventricular fibrillation (VF), Atrial fibrillation (A FIB) and supra ventricular tachycardia (SVT). An essential part of this research was the development of a portable high resolution ECG device, capable of connecting with, either an ECG simulator device, or recording real human data. This device is able to produce higher resolution than normal ECG devices and high values of Signal to Noise Ratio (SNR). Matlab was used to develop a program that could further examine, analyze and study the ECG samples. Since the heart waveform can be simulated by cubic spline interpolation, this feature was used by the implemented Matlab program. The ECG samples were normalized and processed to produce 4 specific coefficients. These 4 coefficients of cubic spline were used in the applied methodology in order to evaluate and separate the various heart disorders with mathematical terms and equations. The database created was compared with the real human samples that were taken and passed through the same data process. Through this step, the entire data process and implementation was not only confirmed, but also proved that the capability to diagnose heart disorders was possible. Based on the results of the applied methodology, the categorization of heart disorders without actual clinical examination is possible. Further analysis of each group of results, can lead to heart disorder prediction. Also given are further suggestions to plan experiments for future work.

The 3rd Generation Partnership Project (3GPP) Long Term Evolution (LTE) is the potential key to meet the exponentially increasing demand of the mobile end users. The entire LTE network architecture and signal processing is carried out at the enhanced NodeB (eNB) level, hence the increased complexity and cost. Therefore, it is not efficient to deploy eNB for the purpose of extending the network coverage. As a solution, deployment of relay node (RN), with radio-over-fibre (RoF) acting as the interface between eNB and RN is proposed. Due to the high path loss and multipath fading, wireless interface would not be the ideal channel between eNB and RN. A detailed investigation is carried out by comparing the Rayleigh multipath fading channel with the optical fibre channel, where the latter achieved a ~31 dB of signal-to-noise ratio (SNR) gain. The distributed feedback laser (DFB) is selected as the direct modulated laser (DML) source, where the modulation method introduces a positive frequency chirp (PFC). The existing mathematical expression does not precisely explain on how the rate equations contribute to PFC. Therefore, an expression for PFC is proposed and derived from the carrier and photon densities of the rate equations. Focusing on theoretical development of DML based RoF system, a varying fast Fourier transform (FFT) scheme is introduced into LTE-Advanced (LTE-A) technology as an alternative design to the carrier aggregation. A range of FFT sizes are investigated with different levels of optical launch power (OLP), the optimum OLP has been defined to be within the range of ~-6 to 0 dBm, which is known as the intermixing region. It is found that FFT size-128 provides improved average system efficiency of ~54% and ~65% in comparison to FFT size-64 and FFT size-128, respectively, within the intermixing region. While fixing FFT size to 128, the investigation is diverted to the optimisation of optical modulators. The author revealed that the performance of dual electrode-Mach Zehnder modulator (DE-MZM) is superior to both DML scheme and single electrode (SE)-MZM, where DE-MZM achieved a transmission span of 88 km and 71 km for 16-quadrature amplitude modulation (QAM) and 64-QAM, respectively. At the initial experimental link design and optimisation stage, an optimum modulation region (OMR) is proposed at the optical modulation index (OMI) of 0.38, which resulted in an average error vector magnitude (EVM) of ~1.01% for a 10 km span. The EVM of ~1.01% is further improved by introducing the optimum OLP region at –2 dBm, where the observed average EVM trimmed to ~0.96%. There is no deviation found in the intermixing region by transmitting the LTE signal through a varying transmission span of 10 to 60 km, additionally, it was also revealed that the LTE RoF nonlinear threshold falls above the OLP of 6 dBm. The proposed system was further developed to accommodate 2×2 multiple-input and multiple-output (MIMO) transmission by utilising analogue frequency division multiplexing (FDM) technique. The studies procured that the resulting output quality of signal at 2 GHz and 2.6 GHz is almost identical with a twofold gain in the peak data rate and no occurrence of intermodulation (IMD). In order to emulate the complete LTE RoF solution, an experimental design of full duplex frequency division duplex (FDD) system with dense wavelength division multiplexing (DWDM) architecture is proposed. It is found that channel spacing of 50 MHz between the downlink (DL) and uplink (UL) introduces severe IMD distortion, where an adjacent channel leakage ratio (ACLR) penalty of 14.10 dB is observed. Finally, a novel nonlinear compensation technique utilising a direct modulation based frequency dithering (DMFD) scheme is proposed. The LTE RoF system average SNR gain observed at OLP of 10 dBm for the 50 km transmission span is ~5.97 dB. External modulation based frequency dithering (EMFD) exhibits ~3 dB of average SNR gain over DMFD method.

This thesis is a study of three techniques to improve performance of some standard fore-casting models, application to the energy demand and prices. We focus on forecasting demand and price one-day ahead. First, the wavelet transform was used as a pre-processing procedure with two approaches: multicomponent-forecasts and direct-forecasts. We have empirically compared these approaches and found that the former consistently outperformed the latter. Second, adaptive models were introduced to continuously update model parameters in the testing period by combining ?lters with standard forecasting methods. Among these adaptive models, the adaptive LR-GARCH model was proposed for the fi?rst time in the thesis. Third, with regard to noise distributions of the dependent variables in the forecasting models, we used either Gaussian or Student-t distributions. This thesis proposed a novel algorithm to infer parameters of Student-t noise models. The method is an extension of earlier work for models that are linear in parameters to the non-linear multilayer perceptron. Therefore, the proposed method broadens the range of models that can use a Student-t noise distribution. Because these techniques cannot stand alone, they must be combined with prediction models to improve their performance. We combined these techniques with some standard forecasting models: multilayer perceptron, radial basis functions, linear regression, and linear regression with GARCH. These techniques and forecasting models were applied to two datasets from the UK energy markets: daily electricity demand (which is stationary) and gas forward prices (non-stationary). The results showed that these techniques provided good improvement to prediction performance.

The primary goal of this research work is to investigate the use of slot loading in reduced size planar fed microstrip patch antennas and develop new antenna structures based on this technique. At present, little theoretical investigation or design methodology exists to support the design of compact structures and research in this field is largely empirical. Moreover, little work exists on the use of planar fed designs. This necessitates a primary requirement to firstly address this knowledge gap. To facilitate this, a mathematical modelling technique that can be applied to such structures is developed. This is based upon the segmentation and Green's function approach. Using this model, the performance of slot loaded structures in terms of circuit characteristics including resonant frequency, input impedance, and Q factor is determined. Using this knowledge, a design procedure is established and subsequently used to provide a framework for the design of novel slot loaded antennas for specific applications. Several new slot loaded patch antenna configurations are designed that produce size reduction whilst allowing the use of a planar feed. The validity of the designs are confirmed through the use of commercial full-wave modelling software package Ensemble. Three linear polarised antennas are presented which are shown to achieve size reduction of 12, 40 and 55% respectively. Several compact circular polarised antenna structures are successfully implemented producing size reduction of up to 43%. A novel design for a reduced size antenna with a dual frequency response is also presented with a tuneable frequency ratio of between 1.03 — 2.0. Prototypes of the aforementioned antennas are fabricated and tested, and practical results are shown.

This research project aims at evaluating a conversion system based on the emerging Brushless Doubly Fed Reluctance Machine (BDFRM) through a comparative experimental study with a traditional and well established slip-ring counterpart, the Doubly Fed Induction Machine (DFIM). One of the main objectives is to establish whether this alternative machine is worthy of industrial consideration in variable speed applications with limited speed ranges (e.g. wind turbines, pump-like drives etc.) in terms of control, reliability, efficiency and power factor performance as major criteria. Such kind of work has not been reported in the open-literature to date and represents the main contribution of the project being undertaken. A conventional and widely used parameter-independent vector control (VC) scheme has been selected for the operation of both the machines using a shaft-position sensor. The VC algorithm has been simulated and implemented in real-time on state-of-the-art eZdsp development platform based on the TMS320F28335 Digital Signal Controller (DSC). The control code has been derived from a programme written in C++ using the corresponding compiler, the Code Composer Studio (CCS). Comprehensive computer simulations have been done in Matlab/Simulink using the parameters obtained by off-line testing of the DFIM and BDFRM prototypes, which have been built in the same stator frame for comparison purposes. The simulation results have been experimentally verified on two identical test rigs where a commercial 4-quadrant cage induction machine V/f drive has been used as a prime mover or load for either the DFIM or the BDFRM subject to their operating mode. The preliminary experimental results on two small-scale prototypes have shown that the BDFRM can achieve competitive performance to the similarly rated DFIM and as such should warrant further investigation and increasing interests of both academic and industrial communities as a potential large-scale wind generator or a pump drive.

In wireless communications and radar systems, there are requirements for high efficiency, small size, low cost, and wide bandwidth of transmitter front –end usage for commercial and also military applications. Active integrated antenna (AIA) could satisfy almost all the requirements. The overall objective of the proposed research is to model, optimise, and design a compact and high efficient AIA using an aperture coupled microstrip antenna (ACMA) by integrating with a power amplifier (PA). Research on ACMA has been focused on the transmissions line (TL) model (TLM) and full wave electromagnetic (EM) model analysis. The full wave investigation is rigorous and elegant but because the dimension of the physical model and the value of the circuit elements are interdependent, the design of the antenna is still difficult. TLM analysis has lower accuracy but easier to analysis and optimise than full wave EM model analysis. To increase the accuracy, the challenge is the coupling ratios between feed/slot, and slot/patch where no unique solution at the moment exists. In this thesis, a novel and simplified method has been produced to investigate these ratios using Scattering (S) parameters. A dual frequency ACMA has been designed to verify these results. Research on the class F and inverse class F PAs is carried out by a novel and simplified load/pull method. A new design method of harmonic load matching network has been presented using lump elements and TLs. Both linear and nonlinear modelling has been investigated. High power added efficiency (PAE) and high gain which are up to 60% and 12dB have been obtained. Finally AIAs have been produced based on previous investigation on class F, inverse class F PAs and a broadband circular polarized ACMA design with 350 MHz bandwidth and 8.5 dB gain at 2 GHz.

This thesis proposes to marry two separate technologies together. The first technology is that of visible light communications (VLC), and the second is small molecule and polymer organic photonic devices. These two technologies both offer outstanding potential in their respective fields of information communications and optoelectronics, with both being proposed as two of the most important technologies about to emerge in the next decades by their respective research communities. As such, it is imperative to investigate and analyse the performance of organic photonic devices in the context of VLC broadcasting networks. There have been no experimental results in the literature reporting on organic-VLC systems until the work proposed in this thesis and therefore the focus is on improving transmission speeds.

The brushless doubly fed reluctance machine (BDFRM) is an attractive alternative technology to traditional slip-ring doubly fed induction machine (DFIM) for slip-power recovery applications with limited variable speed ranges such as wind turbines or pump-alike drives. Owing to its favourable operational characteristics, the BDFRM has been receiving increasing attention by the research and industrial communities, foremost because of its brushless design and high reliability (these being the main DFIM limitations) as well as cost advantages of using a partially rated power electronics converter. A comprehensive, largely or small scale lab model machine parameter independent control development and comparative performance analysis of this emerging machine topology have been done both by computer simulations and experimentally on a dSPACE platform using the parameters of a small-scale laboratory prototype obtained by off-line testing. A Maximum Torque Per Inverter Ampere (MTPIA) strategy and scalar control method with voltage boost, which is rather suited to drive and generator systems under consideration where fast dynamic response is not required, have been implemented and efficiency of the machine (e.g. copper losses) investigated in both motoring and generating modes. Such kind of work has not been reported in the literature available on the subject. A large-scale BDFRG for grid-connected wind turbines with maximum power point tracking (MPPT) has also been examined by simulation studies under the same scalar control conditions referenced above. Comparisons with high-performance control algorithms as primary flux (field) oriented control (FOC) have been made and associated trade-offs considered for typical ii wind profiles. A vector control (VC) scheme, similar to the FOC approach, has also been demonstrated on both the machine side converter (MSC) and grid side converter (GSC) in the BDFRM(motoring and generating)modes. The BDFRG low-voltage fault-ride through capability has also been researched and a dedicated FOC scheme developed to allow the generator to stay on-line and provide the necessary reactive power support for the network voltage recovery under the faulty operating conditions. The BDFRG has been shown to be able to safely ride through the fault without the crowbar circuitry this being difficult to achieve with a comparable DFIG. This desirable BDFRG property can be attributed to the relatively large leakage reactance’s and consequently lower fault current levels compared to the equivalent DFIG.

The main objective of this research is to design and implement an eight-hydrophone transmitter array for generating bipolar acoustic pulses mimicking those produced by cosmogenic neutrino interaction in sea water. In addition, the research was conducted as part of the ACoRNE collaboration. The work initially investigated a single hydrophone system. Due to the nature of hydrophone, the acoustic output signal does not precisely follow a given driving voltage input. Hence signal processing techniques and hydrophone modelling were applied. A bipolar acoustic generation module was built using 8-bit PIC microcontrollers for processing and control. A NI USB-6211 National Instruments commercial module was used for validation of results. The modelling was compared to experimental data generated in a water tank, showing excellent agreement. This single hydrophone instrument was deployed at the Rona array in 2008. Both 10 kHz and 23 kHz pulses were injected, whilst seven hydrophones at Rona site were chosen as the receiver hydrophone array. Signal processing techniques were applied to identify these pulses. The result showed that the triggered pulses can be detected and identified at Rona over a distance of a few hundred metres. A model for an eight-hydrophone transmission linear array system for the ANTARES site was developed. The simulation showed that the eight hydrophones arranged over an eight-metre spacing structure can mimic the anticipated pancake behaviour predicted from neutrino-induced showers as well as generating the acoustic bipolar pulse shape of sufficient amplitude for detection at ANTARES. An eight-channel arbitrary waveform generator module was designed and built using 16-bit dsPIC microcontrollers. Signal Processing techniques were again applied to calibrate the hydrophone transmitter array. The behaviour of an acoustic transducer array was examined in a laboratory water tank to study the shape and direction of such a signal in water. The results were validated against a PXI-6713 commercial module. Excellent agreement was achieved. Finally, the system was deployed at the ANTARES site in September 2011. A range of test signals including 23 kHz bipolar pulses, sine signals and orthogonal signals were injected into seawater to simulate neutrino interactions and investigate signal coding. Signal processing techniques were applied to the data deployed in order to recognise the signals emitted. However, the vessel was far away from the position planned (c 1km), hence the signal received was too weak and no signal was detected. However, the deployed data is still very useful in order to study the noise background of seawater and much has been learned for future sea campaigns.