This thesis describes research work that the author has undertaken and published in the field of electronic reliability prediction techniques over the last 25 years. Reliability prediction is an important area since it has been part of the backbone of reliability engineering in one form or another for over fifty years. The author has over 45 publications that are within the area of reliability prediction and 13 of these have been selected for review in this thesis. In order to show how the author’s work has contributed to the field of reliability prediction this document also contains information on the history of reliability prediction. This allows the author’s work to be placed in context with general developments in the field. The contributions to knowledge and innovations that have been made in reliability prediction include the development of statistical models for lifetime prediction using early life data (i.e. prognostics); the use of non-constant failure rates for reliability prediction; the use of neural networks for reliability prediction, the use of artificial intelligence systems to support reliability engineers’ decision making; the use of a holistic approach to reliability; the use of complex discrete events simulation to model equipment availability; demonstration of the weaknesses of classical reliability prediction; an understanding of the basic behaviour of no fault founds; the development of a parametric drift model; identification of the use of a reliability database to improve the reliability of systems; and an understanding of the issues that surround the use of new reliability metrics in the aerospace industry.

A number of algorithms have been developed in the past for the purposes of target tracking, these have generally been for simple polygonal environments. However as the technology for autonomous vehicles develops for use in the real world these tracking algorithms need to be tested in larger more realistic environments. This work investigates the use of tracking algorithms to control a team of road based robotic platforms, tracking pedestrian targets in urban environments. The novelty of this work is in the identification of the aspects of the environment that affect target tracking algorithms, and modifying the algorithms to cope with them. Problems such as the frequent stalemates reached as an algorithms movement is limited by the highly restricted movement space or the identification of “short cuts” in which the target can take much shorter routes between positions than the robots. Algorithms are developed that overcome these limitations and they are tested in a simulation that is an accurate representation of a real environment. The algorithms are partly based on existing work and are developed extensively to be suitable for the environment. These algorithms are tested for their ability to maintain visual contact with the target. The scenario is tested with varying numbers of robots, speeds and locations. Three algorithms were developed and tested, one built as an extension of existing target tracking algorithms (Combined Urban Tracker) and another two algorithms developed specifically for this environment (Short Cut Path, and Branch). It is concluded that the Combined Urban Tracker and Short Cut Path algorithms performed comparably with a less than 0:3% difference in performance between the two both averaging roughly 54% effectiveness overall, however the Branch algorithm fared significantly worse averaging only 43% overall. The areas within the environment that give significant problems are large open spaces and areas that are significantly occluded from the road network. This work provides a platform on which further development in this area can be based in order to progress tracking algorithms towards being of practical use.

Photovoltaic Concentrator systems, which increase the solar radiation intensity on the photovoltaic cells, may reduce the system cost, if the cost of the concentrator is less than the photovoltaic material displaced. An Asymmetric Compound Parabolic Photovoltaic Concentrator (ACPPVC) for building façade integration with a solar concentration ratio of 2.0 has been designed, fabricated and experimentally characterised. The truncated ACPPVC has acceptance half angles of 0° and 55° and an absorber width of 125mm. Phase Change Materials (PCM) have been integrated to the rear of the PV panel to moderate the temperature rise of the PV and maintain good solar-electrical conversion efficiency. The thermal behaviour of a Fresnel lens PV Concentrator (FPVC) has also been studied in this work. A two-dimensional ray trace technique has been used to predict the optical performance and the angular acceptance of the ACPPVC system. The predicted highest optical efficiency was 88.67% for the ACPPVC-55 system. Extensive indoor experimental characterisation of a number of PV systems was undertaken for a range of incident solar radiation intensities using a highly collimated solar simulator developed specifically for this project. Experimental results showed that the electrical output from the ACPPVC-55 was approximately 1.8 of that of a non-concentrating PV system with similar solar cells area. The electrical conversion efficiency for the ACPPVC-55 system was further increased, when RT27 PCM was incorporated to its rear.

Power converters enable e±cient conversion of electric power, thereby reducing power consumption and cost. The key enabling technologies inside power converters, typically used in hybrid electric vehicles, are the semiconductor devices. Device reliability is of high priority because they generate heat from the dissipation of electric power which can lead to failure if the device maximum junction temperature is exceeded. Furthermore, device temperatures can vary largely in switching applications, leading to thermal-mechanical fatigue failure. As electronic designers are pushed to deliver smaller and more powerful packages, they are finding thermal issues increasingly di±cult to solve. The primary goal of this thesis is to develop a fast and accurate thermal simulation design tool which is capable of simulating realistic power converter operation. Most commercial thermal simulators use finite-element software. Despite their perceived accuracy, they suffer from severe computational requirements and offer limited ability to explore power converter packaging converter designs during realistic converter operation. Traditional approaches using R-C networks as thermal equivalent circuits are of little use as a design tool since for every geometrical layout of the packaging structure which is tested, the designer must return to the starting point which is either a time-consuming FE simulation or a practical experiment. The Fourier thermal model presented in this thesis is a purely conductive model requiring no parameter extraction or use of a FE simulator. The starting point for the Fourier model is the heat equation. The Fourier thermal model yields solutions to the heat equation by carrying out spatial discretisation using a truncated Fourier series and using MATLAB/Simulink to perform temporal discretisation using a dynamic ODE solver. Validation using the finite volume thermal simulator FLOTHERM showed that the transient Fourier model could accurately simulate 3-D heat conduction through a wide range of power converter packaging structures. The Fourier thermal model is an excellent early stage design tool because its simulation speed is far superior to FLOTHERM, even though both models operate with a similar accuracy. The use of MATLAB/Simulink as the simulation environment enabled the Fourier thermal model to operate within the framework of an electro-thermal simulator and therefore simulate realistic load conditions. This is major advantage over existing approaches which fail to simulate electro-thermal interaction. Experimental validation of the fast electro-thermal converter simulator was achieved by utilising an inverter back-to-back rig and recording transient device temperatures using an infrared camera. The similarity between the experimental and simulated results indicated that the Fourier thermal model was sufficiently accurate. The electro-thermal simulator operated at a simulation speed which was ten times real time, which is extremely fast compared to existing approaches which can take up to two days to simulate a 60 second drive cycle. 'Ten times real time' represents a significant step forward for power converter packaging design. In the future device reliability can be accurately predicted if the electro-thermal simulator model is combined with a reliability model. The potential of the Fourier thermal model to aid numerical optimisation of the whole power converter is exciting.

The primary aim of this research project was to deposist thick-film (low-frequency) s-mode CdS piezoelectric transducers directly onto copper or aluminium rods which formed part of the welding-electrode of a spot-welding machine. These transducers were to replace the discrete transducers used in a set-up for the "on-machine" evaluation of spot-welds. A secondary aim was to deposit very thin CdS films on glass slides for use as microwave resonators. The dependence of film adhesion on film thickness, and crystallographic orientation changes with thickness imposed an upper limit on the thickness of transducers with adequate properties for applications. The final goal, established through experience on the project, was to determine how thick piezoelectric films could be deposited to make useful transducers. Highly stoichiometric (deviations from stoichiometry of the order of 1 part in 10^13 ) and highly resistive (> 10^10 Ω.m.), and highly oriented films up to 100 μm thick have been successfully deposited on Ae rod substrates. Two deposition techniques were used : CdS/S electron beam bombardment evaporation and Cd/S isothermal cells. Provided that the temperature of the vapour molecules was less than 400 degrees C and that the pumping speed could be increased at will , then, the faster the deposition rate, the sharper the oblique c-axis preferred orientation, and the better the piezoelectric performance of the films. The pumping speed limited the deposition rate to 10 μm.h^- 1. Appreciable thermal stresses in the films gave rise to large forces which induced the thick-films to flake off or disintegrate. The dependence of film adhesion on film thickness is explained in terms of the inequality between the forces which bind the film to the substrate (independent of thickness) and the forces which induce the film to flake off (proportional to thickness). Thick CdS films were made to adhere to the substrate by making the substrate surface rouqher so that the films "keyed-in". No appreciable temperature gradients existed in the CdS films during growth, either across their thicknesses or along their surfaces. No changes in temperature gradients occurred in the films due to changes in film orientation, and vice versa. Up to a certain critical thickness, the c-axes of most CdS film crystallites aligned themselves with the direction of the vapour beam. When the thickness of the film exceeded the critical thickness, the growth of oblique crystallites was stifled and the film's c-axis tilted towards the substrate-normal and eventually became parallel to it. This was confirmed by etching-back a thick CdS film which was deposited at oblique vapour incidence. A model is presented for the "stifling process” which gives the relation between the critical thickness, the grain size and the deposition angle of the film. For a given deposition environment, the stifling process imposed an upper limit on the thickness of an s-mode transducer. The use of copper substrates, and of copper parts inside the deposition chamber, was abandoned because of the corrosive action of sulphur on copper. Cu/CdS junctions were nearly ohmic, and the anomalous behaviour of these junctions is explained in terms of the reaction between Cu and S to form Cu2S. CdS s-mode transducers with untuned two-way insertion loss of 35 dB in a 50 ohm system have been successfully deposited on glass slides for operation at frequencies down to 20 MHz. The stress in CdS films on glass slides was much less than that on Ae rods. It is possible that the higher stress in films on Ae rods weakened their piezoelectric performance.

Air-coupled thin-membrane capacitive ultrasonic transducers have been developed that use microstereolithography fabrication with architectures comprised entirely of partially metalised photopolymer. These devices derive considerable advantages from rapid prototyping technology, in that they are cheap to produce, and benefit from the design-to-product lead times inherent in the production of components using stereolithography. To date membranes have been produced with thicknesses ranging from 30 to 90 μm with aspect ratios in the range of 100 - 1000. These devices have been shown to operate both as transmitters and as receivers of ultrasound, and have a bandwidth approaching 100% with a centre frequency of 100 kHz. The method of fabricating these devices allows for easy modification for various applications including structural health monitoring and immersion, as well as affording the possibility of integrated focussing or wave-guiding architecture and packaging that can be incorporated into a single build. Fundamental or subtle changes to a given transducer design may be achieved incurring little additional cost or time. This novel approach to transducer fabrication enables the bespoke manufacture of specific transducer architectures from a computer aided design package using polymers that exhibit different material properties to materials used in silicon micromachining, but at the same time allow for fabrication on a scale that is approaching that of microfabrication. The versatility of 3-D rapid prototyping allows the realisation of more complicated structures than was possible previously. This work examines these transducers in terms of their characterisation and their operation in conjunction with other transduction architecture, such as focussing parabolic mirrors that were also produced using the same manufacturing technology. In addition, their operation in contacting acoustics and the reception of surface acoustic waves has been demonstrated. Immersion studies using these devices have found that that they hold promise for operation in a variety of different media. These transducers are seen as an important prototype development tool in the field of capacitive ultrasonic transduction and microphone-speaker design.

The research work described in this thesis concentrates on the application and feasibility of FACTS controllers as a solution to the problem of congestion management. Effective congestion management has become increasingly important since wholesale electricity markets became privatised and competition was encouraged between generation and load (supply) companies. In addition, growth in load demand, planning permission and the need to integrate renewable generation sources will push transmission systems to work closer to their operating limits. Therefore, ensuring future systems are able to sustain these new conditions is a considerable challenge. The first part of the work focuses on the integration of the bilateral market and steady state FACTS controller models into the interior point optimal power flow (lP OPF) algorithm. The objective is to minimise changes to scheduled bilateral market contracts to provide economic savings for the transmission system operator (TSO) within congested systems. The optimisation technique is able to reduce the level of congestion. Early results identify that congestion is the dominant proportion of the system costs incurred while system losses are of similar levels with and without controller installation. A functional procedure is proposed for fair and easy comparison and a general two-step method is introduced which aims to find optimal location and rating of the installed controller by applying the IP OPF algorithm. In the latter part, the general two-step method is utilised to assess the performance of FACTS controllers in terms of financial benefits over an average year, and considers daily and seasonal changes in demand.I n practice, transmission systems contain thousands of buses therefore a sensitivity-based three-step method is developed to reduce the number of required simulations to find the optimal location and rating of FACIS controllers. The method has been successfully demonstrated on IEEE standard test systems and has the potential to act as a first-step screening technique for practical systems. The last section of this work is concerned with assessing the economic feasibility of FACTS controllers as a congestion management solution by introducing an economic measure - a return index. This measure is used to assess the viability of different locations as it compares the relative cost savings of the TSO to the equipment cost of the controller. In conclusion, the results show that with appropriate location choice, installation of specified FACTS controllers are able to provide a solution to the congestion management problem with realistic Payback periods.

This thesis describes the development of a sensorless control method for a surface mounted permanent magnet synchronous machine drive system. The saturation saliency in the machine is tracked from the stator current transient response to the fundamental space vector PWM (pulse width modulation) excitation. The rotor position and speed signals are obtained from measurements of the stator current derivative during the voltage vectors contained in the normal fundamental PWM sequence. In principle, this scheme can work over a wide speed range. However, the accuracy of the current derivative-measurements made during narrow voltage vectors reduces. This is because high frequency current oscillations exist after each vector switching instant, and these take a finite time to die down. Therefore, in this thesis, vector extension and compensation schemes are proposed which ensure correct current derivative measurements are made, even during narrow voltage vectors, so that any induced additional current distortion is kept to a minimum. The causes of the high frequency switching oscillations in the AC drive system are investigated and several approaches are developed to reduce the impact of these oscillations. These include the development of a novel modification to the IGBT gate drive circuit to reduce the requirement for PWM vector extension. Further improvements are made by modifications to the current derivative sensor design together with their associated signal processing circuits. In order to eliminate other harmonic disturbances and the high frequency noise appearing in the estimated position signals, an adaptive disturbance identifier and a tracking observer are incorporated to improve the position and speed signals. Experimental results show that the final sensorless control system can achieve excellent speed and position control performance.

In this thesis computational studies on electromagnetic properties of metamaterials and applications are introduced. The studies include an introduction to the time-domain transmission line modelling (TLM) method and the fundamental scattering properties of metamaterials. The first major objective of the thesis is directed to predicting the resonant frequencies of some forms of cut-wire (CW) metamaterials by using approximate equivalent circuits. The second objective is the improvement of metamaterial simulation efficiency based on two approaches: a simulation method based on retrieved metamaterial electromagnetic properties and one based on digital filtering (DF) techniques. By using the DF techniques, metamaterials are effectively modelled and the simulation times are significantly reduced. The third objective is focused on studying CW metamaterial as potential absorbers by deliberately including conductive losses. The proposed CW metamaterials are found to exhibit customisable absorptance characteristics with arbitrary polarisation. This metamaterial absorber study includes an experimental validation.

Combined Heat and Power Generation (CHP) or cogeneration has been considered worldwide as the major alternative to traditional energy systems in terms of signi ticant energy saving and environmental conservation. A renewable energy resource-fuelled CHP would deliver even more environmental benefits than a fossil tuel-driven CHP. Biomass is one of the renewable energy resources that plays an important role to the world primary energy supplies and can be used to fuel CHP systems. Many medium- and large-scale biomass-fired CHP plants have been demonstrated and commercialized in the world. However, biomass-fuelled microscale CHP (1-1 OkW c>) which is suitable tor building applications has yet to be commercialised or demonstrated. The development and evaluation of a micro-CHP system operating on biomass energy has been the focus of this PhD research. It is an integral part of an externally funded research project which aims to develop and evaluate a novel, first-of-its-kind, micro-scale (l - 2 kWe) biomass-tired CHP system suitable for public and large domestic buildings' applications. The specific tasks of the present PhD research arc: To thermodynamically model the micro-scale biomass-fired CHP system with organic Rankine cycle (ORC): different environment-friendly working fluids are to be modelled with the ORC processes. To experimentally evaluate the micro-scale biomass-tired CHP system in terms of power generation and combined heat and power pertormance. To experimentally investigate the combustion performance and NOx emissions of the biomass pellet boiler which is a key component of the micro-scale biomass-fired CHP system. The micro-scale biomass-tired CHP system with ORC developed by the research team of University of Nottingham including the author of this PhD thesis mainly consists of a biomass boiler, an ORC fluid evaporator, an ORC turbine, an alternator, a heat reeouperator and a condenser. The boiler produces hot water which transfers heat to the organic working tluid via the evaporator. The generated organic fluid vapour drives a turbine to rotate an alternator, producing power. The expanded organic fluid vapour leaving the turbine transfers some of its heat to the recouperator and then is condensed by cooling water which can be heated to around 40 - 50 °C for domestic washing and under-floor heating purposes. The main methodologies of the present PhD research are the thermodynamic modelling of the proposed micro-scale biomass-tired CHP system with ORC and the laboratory testing of the assembled micro-scale biomass-fired CHP system and its main components (biomass boiler, ORC turbine, alternator, heat exchangers etc.). Literature review has demonstrated that the biomass-fired micro-CHP systems for buildings present many advantages compared to conventional separate heating and power supply systems (e.g. a dedicated boiler for heating and grid for power supply) as they can present higher primary energy savings and lower CO2 emissions. ORC is a suitable thermodynamic cycle that could be used for micro-CHP systems while operating with waste heat and renewable energy resources which are available at relative low temperatures. Thermodynamical modelling of the proposed micro-scale biomass-fired CHP system with ORC has been carried out and the results have been presented and discussed in the thesis. Three different environment-friendly working fluids, namely HFE7000, HFE7100 and n-pentane, have been modelled with various ORC process configurations. The laboratory testing of the assembled micro-scale biomass-fired CHP system and its main components (biomass boiler, ORC turbine, alternator, heat exchangers etc.) has been carried out initially with a 25kW biomass boiler and then with a 50kW biomass boiler. The main purpose of the laboratory testing has been to evaluate the main energy efficiencies (the electrical efficiency and the total CHP efficiency) of the assembled micro-CHP systems. The combustion performance and NOx emissions of each biomass boiler have also been investigated as the biomass boiler is a key component of the micro-scale biomass-tired CHP system. The experimental findings of these laboratory tests are presented and analysed in the thesis. Finally, the conclusions of the present PhD research have been given. The modelling results have shown that the electrical efficiency of the micro-CHP system depends on not only the modelling conditions but also the ORC fluid. A comparison of the three fluids generally follows the following order: n-pentane > HFE7000 > HFE7100. For the laboratory test, the 25kW biomass boiler-driven micro-CHP system, having an ORC efficiency in the range of 2.20% - 2.85%, can generate electricity of 344.6W and heat of 20.3kW, corresponding to electricity generation efficiency 1.17% and CHP efficiency 86.22%. And the 50kWth biomass boiler-driven micro-CHP system, having an ORC efficiency of 3.48% - 3.89%, can generate electricity of 748.6W and heat of 43.7kW, corresponding to electricity generation efficiency 1.43% and CHP et1iciency 81.06%.