Master of Science / Department of Electrical and Computer Engineering / William B. Kuhn / Q‐enhanced filtering technologies have been heavily researched, but have not yet been adopted into commercial designs due to tuning complexity, and performance issues such as noise figure and dynamic range. A multi‐pole Q‐enhanced band‐pass filter operating at 450 MHz with tunable bandwidth is developed in this thesis. A noise figure of 14 dB and dynamic range of 140 dB/Hz have been measured, making the filter suitable for operating in the IF subsystem of a radio receiver. The design utilizes off‐chip resonators, created using surface mount components or embedded passives in LTCC processes, to have a reasonably high base‐Q. The equivalent parallel loss resistance of the finite‐Q inductor and connected circuitry at resonance is partially offset by negative resistances, implemented with tunable on‐chip transconductors, as required to reach the needed Q for the targeted bandwidth. Each pole of the filter has binary weighted negative resistance cells for Q‐enhancement and binary weighted capacitances for frequency tuning. Binary weighted capacitive coupling cells allow the filter to achieve the level of coupling appropriate to the targeted bandwidth. To maintain the filter bandwidth, center frequency, and gain over environmental changes a realtime tuning algorithm is needed. A low complexity tuning algorithm has been implemented and found to accurately maintain the bandwidth, center frequency, and gain when operating at bandwidths of 10 or 20 MHz. Flatness of the pass‐band is also maintained, to within 0.5 dB across a temperature range of 25‐55 degrees C. In addition to the implementation of the tuning algorithm, the thesis provides a solution for pass‐band asymmetries spawned from the use of finite‐Q resonators and associated control circuitry.

The objective of this research is to show that nonlinearity can be used to improve vibration absorption and suppression of unwanted vibrations in a main system due to external excitation. This was shown by investigating two systems a SDOF (with hardening nonlinearity) and a 2DOF (with softening nonlinearity). The aim of carrying out these investigations was to introduce a passive nonlinear system that can update itself and self-regulate to suppress undesired oscillations. To fulfil the desired gaol, various types of springs were considered and investigated. A commercially available spring called Mag-spring has been chosen and a function for its nonlinearity has been investigated. Mag-Spring is a newly invented spring which is designed to exhibit constant force at its operating range. However, this spring has a special non-linear behaviour before reaching to the constant force domain which is the main focus of the investigation presented in this thesis. The nonlinear behaviour of Mag-spring encouraged the idea that vibration design is possible by the advantages that can be gained from magnetic technology. The added benefit through this new Mag-spring, is that it solved some of the concerns assotiated with old available ones. The most concern norrowing the usage of magnetic springs, is the air gap between the two magnets, which make the attraction or repulsive force unstable through the spring’s working range. Linmot Company, introduced a solution to this concern by introducing a teflone that works as a bearing between the two magnets, which fixed the distance between them while they are sliding againest each other. In the first scenario of this study, a hardening nonlinear spring was added in parallel to a system with a single degree of freedom. The system will remain single degree of freedom as the spring was added in parallel without additional mass. The hardening spring shows low stiffness at low amplitude and high stiffness at higher amplitude. In this study, it was shown that nonlinearity affects the dynamic performance of a system and makes the natural frequency amplitude dependant. As the amplitude of vibration increases, consequently, stiffness increases and the natural frequency shifts away from the excitation frequency. For this investigation, a vibrating system with one degree of freedom has been built based on a mathematical model simulated and tested in Matlab software. Mag-spring was used to introduce the nonlinear stiffness to the system. Unbalance mass mounted to a disc fixed to a rotational machine has been used to create a forced vibration system with variable frequency. The response of the system with and without nonlinearity effect was monitored with an accelerometer. Simulation and experimental results showed that nonlinearity could shift the resonance frequency of the SDOF system by 10% (hardening of the system), without affecting the stiffness of the system at normal working condition. In the second scenario, a softening nonlinear spring was added as a vibration absorber to a system with a single degree of freedom, to make the system with two degree of freedom. The softening spring shows high stiffness at low amplitude and low stiffness at high amplitude. The rationale behind this is to introduce a spring which is hard at high frequency and soft at low frequency, which as a result will make the ratio √(k_a/m_a ) of the absorber follows the excitation frequency (ω) allowing the system to update itself and self-regulate providing vibration cancellation at more than one frequency value and widen the vibration cancellation range (ω_n2-ω_n1). It was shown that the Mag-spring could show a softening behaviour in a limited domain if its operating position is shifted. A program has been written to simulate the behaviour of all nonlinear system with two degree of freedom (nonlinear absorber). At this program, the maximum amplitude of each time domain was used to produce the frequency domain of the amplitude of the system. The amplitude of the vibration for a linear and a nonlinear absorber was compared. The results showed that the nonlinear absorber suppresses and reduces the vibration amplitude of the main system better than the linear absorbers with up to 60% reduction in magnification ratio and from 5% to 10% in widening the cancellation range (ω_n2-ω_n1). In the last scenario of this study, 4 different ideal softening stiffness curves were introduced based on theoretical methods. Their vibration response was calculated and compared to the nonlinear absorber (Mag-spring) and a linear absorber. This study shows that when nonlinearity is designed properly, it could provide a distinguished vibration cancellation response resulting more than 60% vibration cancellation improvement. This study demonstrated the possibility of developing a passive self-regulating tuned mass system involving the usage of nonlinearity. Nonlinearity will enhance the vibration cancellation by allowing the system to update itself and as a result will make the vibration absorption to be effective within a frequency range rather than single frequency unlike the classical tuned mass system. This study, to the best knowledge of the author, can be classified as an uncommon study in vibration systems investigations.

Lake surface water temperatures (LSWTs) of 246 globally distributed large lakes were derived from Along-Track Scanning Radiometers (ATSR) for the period 1991 to 2011. These LSWTs, derived in a systematic manner, presents an ideal opportunity to study LSWT behaviour on a global scale. In this thesis, the annual cycles of lake-mean LSWTs derived from these data quantify the responses of large lakes’ surface temperatures to the annual cycle of forcing by solar radiation and the ambient air temperature. Minimum monthly net surface solar irradiance (netSSI) strongly influences minimum LSWTs of non-seasonally ice covered lakes (where lake-mean LSWT remains above 1 ºC throughout the annual cycle), explaining > 0.88 (R2 adj) of the inter-lake variation in both hemispheres. In some regions, for seasonally ice covered lakes (where lake-mean LSWT remains below 1 ºC for part of the annual cycle) the minimum monthly netSSI is a better predictor than latitude, of the length of the frozen period, which shows the importance of local cloud climatological conditions. Additionally, at lake locations between 1º S to 12º N, the netSSI, shown to peak twice annually, is reflected in the LSWT annual cycle. The summer maximum LSWTs of lakes from 25º S to 35º N show a linear decrease with increasing altitude; -3.76 + 0.17 ºC km-1 (R2 adj = 0.95), marginally lower than the corresponding air temperature -4.15 + 0.24 ºC km-1 (R2 adj = 0.95) decrease with altitude. The start and end of the period where the lake-mean LSWT is greater than 4 ºC shows strong correlation with the spring and autumn 0 ºC air temperature crossing days, (R2 adj = 0.74 and 0.80 respectively). The temporally and spatially resolved LSWT observations allows for a greater practical understanding of LSWT behaviour of large lakes. For example, lakes with a greater LSWT annual range have more observed variability in the LSWT extremes, highlighting that they may be more responsive to changes in the climate than lakes with a low annual range. The nighttime LSWT trends show stronger warming than day-night trends in the all regions, except Europe. The lake centre LSWT trends and absolute values can be generally considered representative of the lake-mean LSWT trends and absolute values. The observed LSWT time series are used to tune a 1-dimensional thermodynamic lake model, FLake. By tuning FLake using only 3 basic lake properties, shown by myself to have the most influence over LSWTs (depth, snow and ice albedo and light extinction co-efficient), the daily mean absolute differences for 244 lakes is reduced from 3.38 + 2.74 ºC (untuned model) to 0.85 + 0.61 ºC (tuned model). The effect of wind speed, lake depth, albedo and light extinction co-efficient on LSWTs is demonstrated throughout the tuning process. The modelled summer LSWT response to changes in ice-off is strongly affected by lake depth and latitude explaining 0.50 (R2 adj, p = 0.001) of the inter-lake variation in summer LSWTs. Lake depth alone explains 0.35 (p = 0.003) of the variation, highlighting the sensitivity of the summer LSWTs of deeper lakes to changes in the ice-off. Statistically significant summer/ maximum month modelled LSWT trends, from 1979-2011 are presented for lakes where the modelled LSWTs are strongly supported by observed LSWTs over the period of available observed LSWTs. For these lakes, the trends show LSWT warming of between 0.73 – 2.10 ºC for 29 lakes in northern temperate regions over the 33 year period (1979 – 2011). The modelled regional trends of all lakes over the same period show least warming in Africa of 0.30 ºC and the greatest warming in Europe, 1.35 º.

Motion in the outside world forms one of the primary uses of visual information for many animals. The ability to interpret motion quickly and accurately permits interaction with and response to events in the outside world. While much is known about some aspects of motion perception, there is less agreement about how feature selectivity leading to motion perception is actually formed in the convergent and divergent pathways of the visual system. It is even less clear how these classical understandings of motion processing, often driven by artificial stimuli with little resemblance to the outside world, correspond to responses of neurons when using more natural stimuli. In this thesis, we probe these gaps, first by demonstrating that synchronization within the visual thalamus leads to efficient representations of motion (through tuning properties) in primary visual cortex, exploiting precise timing across populations in a unique manner compared to traditional models. We then create a novel “minimally-natural” stimulus with the appearance of an infinite hallway wallpapered with sinusoidal gratings, to probe how such minimally natural features modulate our predictions of neural responses based upon feature tuning properties. Through encoding and decoding models we find that measuring a restricted tuning parameter space limits our ability to capture all response properties but preserves relevant information for decoding. We finish with an exploration of ethologically relevant natural features, perspective and complex motion, and show that even moderate amounts of each feature within or near the classical V1 receptive field changes the neural response from what classical feature tuning would predict and improves stimulus classification tremendously. Together all of these results indicate that capturing information about motion in the outside world through visual stimuli requires a more advanced model of feature selectivity that incorporates parameters based on more complex spatial relationships.

At the Edinburgh University Collection of Historical Instruments (EUCHMI) there are two virginals where the strings are plucked extremely close to 1/4 of their length over a substantial part of the compass. Principles of Acoustics dictate that in such cases the fourth partial of the resulting sound is unlikely to be produced with any appreciable magnitude. Similar situations affecting different partials also occur when the plucking point is located at other fractional parts of the string length, such as 1/3 or 1/5. A database of plucking points, based on museum catalogues, was created and analysed, so that pitch regions are identified where such phenomena are likely to occur for different categories of instruments. The main cue used in harpsichord tuning is related to beat phenomena produced by partials of two simultaneously played notes, whenever there is a small difference between their frequencies. Partial four being relevant for major thirds and perfect fourths, these intervals may result less accurate, or less reliably tuned, in instruments like the above virginals. Historical sources were surveyed for expressions which describe the cues that tuners were supposed to use in identifying the ideal of an interval. Although expressions such as Schwebungen are mostly related with beats, broader meanings are discussed. Hints of the use of alternative, not beat-related cues, are identified. A series of practical tuning experiments was performed in which selected intervals were tuned, the accuracy of the tuning process being assessed from recordings, which were subjected to spectral analysis. Only cues that could be related to historical sources were used. The effects of factors such as pitch, pitch region, instrument, interval type, and particularly the potential absence of relevant partials were investigated, statistical methods playing a substantial role in the research. A logbook was created, where the experimenter detailed the cues used for each tuned interval in one of the experiments. The recorded information provided some insight into a number of strategies a tuner may use for coping with difficulties in cases of absent or weak partials. The impact of inharmonicity on the exact size of just intervals was also examined, both on theoretical grounds and based on data from the experiments. A number of case studies were included, where a comparison was made between missing partial notes predicted from plucking points and the degree of absence of those partials in the actual spectra for some historical instruments.

This paper presents an approach for an automated calibration process for electronic control units (ECU) of power split transmissions in agricultural tractors. Today the calibration process is done manually on a prototype tractor by experts. In order to reduce development costs the calibration process is shifted from prototype testing to software modelling. Simultaneous optimization methods are used within the software modelling to calculate new parameters. The simultaneous optimization includes objective evaluation methods to evaluate the tractor behaviour. With the combination of both methods inside the software modelling, the calibration process can be automated. The success of this approach depends on the quality of the software modelling. Therefore the identification of the initial prototype behaviour and the fitting of the tractor software model is done at the beginning. At the end of the automated calibration the validation and fine-tuning of the calculated parameters are done on the real tractor. These steps are condensed to a five step automated calibration process which includes simultaneous optimization and objective evaluation methods in several applications. After the detailed discussion of this automated calibration process one function of the ECU (one transmission component) will be calibrated through this process as example.

The objective of this PhD research was to design and implement novel microfluidic radio-frequency (RF) structures on multilayer organic substrates for cooling and tuning purposes. The different designs were implemented to target applications up to C-band (4 GHz – 8 GHz) frequencies. The system-on-package (SoP) solution adopted throughout this work is well adapted for such designs where there is a need to integrate the functionality of different sub-components into a single hybrid fully packaged system. The first part of the thesis is dedicated to the study of a specific liquid cooling scheme using integrated microchannels on organics placed beneath different types of heat sources. A 1 W gallium nitride (GaN) die was cooled using this method and an analysis is presented regarding the cases where the coolant is static or dynamic inside the microchannel. The second part of the thesis deals with microfluidically reconfigurable microstrip RF circuits, mainly bandpass filters and power amplifiers (PAs). The microfluidic tuning technique is based on the change in the effective dielectric constant that the RF signal “sees” when traveling above two microchannels with different fluids. This technique was used to shift the frequency response of an L-band microstrip bandpass filter by replacing DI water with acetone inside a 60 mil micro-machined cavity. This technique was also used to design reconfigurable matching networks which constitute the main part of the proposed tunable GaN-based PA for S- and C-band applications. The final part of the thesis expands the previous results by combining both cooling and tuning in a single RF design. To prove the concept, cooling and tuning microchannels were integrated into a single package to cool a GaN-based PA and tune its frequency response at the same time from 2.4 GHz to 5.8 GHz and vice versa.

As device feature sizes shrink to nano-scale, continuous technology scaling has led to a large increase in parameter variability during semiconductor manufacturing process. According to the source of uncertainty, parameter variations can be classified into three categories: process variations, environmental variations, and temporal variations. All these variation sources exert significant influences on circuit performance, and make it more challenging to characterize parameter variability and achieve robust, low-power designs. The scope of this dissertation is conquering parameter variability and successfully designing efficient yet robust integrated circuit (IC) systems. Previous experiences have indicated that we need to tackle this issue at every design stage of IC chips. In this dissertation, we propose several robust techniques for accurate variability characterization and efficient performance prediction under parameter variations. At pre-silicon verification stage, a robust yield prediction scheme under limited descriptions of parameter uncertainties, a robust circuit performance prediction methodology based on importance of uncertainties, and a robust gate sizing framework by ElasticR estimation model, have been developed. These techniques provide possible solutions to achieve both prediction accuracy and computation efficiency in early design stage. At on-line validation stage, a dynamic workload balancing framework and an on-line self-tuning design methodology have been proposed for application-specific multi-core systems under variability-induced aging effects. These on-line validation techniques are beneficial to alleviate device performance degradation due to parameter variations and extend device lifetime.

Active Queue Management (AQM) aims to achieve high link utilization, low queuing delay and low loss rate in routers. However, it is difficult to adapt AQM parameters to constantly provide desirable transient and steady-state performance under highly dynamic network scenarios. They need to be a trade-off made between queuing delay and utilization. The queue size would become unstable when round-trip time or link capacity increases, or would be unnecessarily large when round-trip time or link capacity decreases. Effective ways of adapting AQM parameters to obtain good performance have remained a critical unsolved problem during the last fifteen years. This thesis firstly investigates existing AQM algorithms and their performance. Based on a previously developed dynamic model of TCP behaviour and a linear feedback model of TCP/RED, Auto-Parameterization RED (AP-RED) is proposed which unveils the mechanism of adapting RED parameters according to measurable network conditions. Another algorithm of Statistical Tuning RED (ST-RED) is developed for systematically tuning four key RED parameters to control the local stability in response to the detected change in the variance of the queue size. Under variable network scenarios like round-trip time, link capacity and traffic load, no manual parameter configuration is needed. The proposed ST-RED can adjust corresponding parameters rapidly to maintain stable performance and keep queuing delay as low as possible. Thus the sensitivity of RED's performance to different network scenarios is removed. This Statistical Tuning algorithm can be applied to a PI controller for AQM and a Statistical Tuning PI (ST-PI) controller is also developed. The implementation of ST-RED and ST-PI is relatively straightforward. Simulation results demonstrate the feasibility of ST-RED and ST-PI and their capabilities to provide desirable transient and steady-state performance under extensively varying network conditions.

Modular multilevel converter type voltage source converter (MMC-VSC) is a potential candidate for present and future HVdc projects. The d-q decoupled control system is widely used to control MMC-VSC systems. Selection of PI-controller parameters for MMC-VSC systems is a challenging task as control loops are not completely decoupled. Since there is no widely accepted method to tune these control loops, the industry practice is to use the trial and error approach that requires a great amount of time. Therefore, it is required to develop a systematic procedure to tune PI-controllers considering necessary system dynamics and also to propose guidelines for control system design. This thesis introduces a systematic procedure to determine PI-controller parameters for the d-q decoupled control system. A linearized state-space model of an MMC-VSC system is developed to calculate the frequency-domain attributes. The control tuning problem is formulated as an optimization problem which is general and any meta-heuristic method can be used to solve the problem. In this thesis, the simulated annealing is applied to solve the problem. The efficacy of the tuned parameters is tested on the electromagnetic transient model of the test system on the real-time digital simulators (RTDS). In addition, it is shown that the proposed method is suitable to tune PI-controller parameters for MMC-VSC systems connected to strong as well as weak ac networks. Further, this thesis investigates the effects of d-q decoupled controller parameters, phase-locked loop (PLL) gains, and measuring delays on the stability and performance of the MMC-VSC test system. It is shown that the converter controllers have greater influence on the system stability and the impact of PLL gains is negligible unless very high PLL gains are used. In addition, the negative impact of measuring delays in instantaneous currents and voltages is also analysed by performing eigenvalue and sensitivity analysis. Finally, a set of guidelines for control design of MMC-VSC systems is summarized. In general, the proposed control tuning procedure would be useful for the industry to tune PI-controllers of MMC-VSC systems. Furthermore, the proposed methodology is generic and can be adapted to tune of any dynamic device in power systems. / February 2017