In this thesis a transformer fault diagnosis system using probabilistic neural network (PNN) is proposed and implemented. Many artificial neural networks (ANN) have been proposed to deal with the transformer fault diagnosis. However, when dissolved gas records change, adaptation capability becomes a problem in ANN applications. PNN analyzes the dissolved gas contents in the oil-immersed transformer to identify various fault types. Numerical gas ratios of oil and cellulose decomposition were used to create the training examples. Retraining can also be done by adding new examples without any iteration. With diagnostic gas records, computer simulations were conducted to show the effectiveness of the proposed system. The Internet based power transformer monitoring system was also proposed in this thesis . LabVIEW was used to develop the Man-Machine Interface (MMI), and DataSocket tool was used to share the information on Internet. Application of the harmonic load flow based on the Equivalent- Current Injection was used to solve harmonic problems. There are two sub-models including the fundamental and harmonic frequency models. The standard Fourier analysis was used to deal with the harmonic loads to get injection currents. A passive filter was also developed to improve harmonics to satisfy restriction standards of the Taiwan Power Company.

Vector Fitting is employed to approximate the Frequency Response Analysis (FRA) measurement data of a 13.8 kV/136.8 kV, 50-MVA single phase transformer. The frequency response of the primary and secondary coils as well the coupling between the coils have been measured in the frequency range of up to 2 MHz. The measured data includes the stray capacitances that exist between the coils. A circuit synthesis model is used to represent the measured data in the form of a two-port passive RLC network. The derived network is implemented in a commonly used power system transient simulator (PSCAD/EMTDC). The proposed RLC model is passive to ensure the stability of the network. The model can be used to investigate transient response of the transformer including the simulation of switching and lightning overvoltage transients. The results are compared with those derived from existing simple models.

In recent years, the trade-off between quality and cost of power system components has become a matter of interest for many utilities. The widespread use of costly electricity networks either in residential or industrial areas has encouraged service providers to find a proper strategy that will minimize the overall life-cycle cost while keeping components in good working condition. The power transformer, which represents approximately 60% of the overall cost of the network, is ranked as one of the most important and expensive components. However, the transformer's sudden failure puts the system in a serious or critical condition which in most cases causes catastrophic loss to both utilities and customers. Significant attention has been given to monitoring and diagnostic techniques that observe any abnormal behaviour, assess the transformer's condition, and therefore minimize the probability of unplanned outage. Yet, applying many various monitoring tests is not always applicable due to the following factors: some tests require the unit to be taken out from service for testing, insufficient availability of man power, and significant cost of applying all the tests. Thus, there is a vital demand for an intelligent method of minimizing the number of monitoring tests without losing much information about the transformer's actual condition. In this research, data mining techniques have been employed to evaluate the transformer's state through intelligent selection criteria that determines the optimal number of monitoring tests in cost-effectiveness. Feature selection technique based on ranker search method has been used to rank the monitoring tests (features) in a priority sequence from their individual evaluation, and to select the most inductive tests that provide the most information about the unit's condition. When the measured data from monitoring tests is collected and prepared, a diagnostic technique is applied to assess the condition of the transformer. In this regard, Support Vector Machine (SVM) has been utilized to perform this task due to its robust classification accuracy. SVM is first applied to the full number of tests, and then the number of monitoring tests is reduced by one after each classification process using the feature selection algorithm. The selected number of monitoring tests has shown the best possible accuracy the classifier can reach over the whole number of tests. Radial Basis Function (RBF) classifier has been used in the classification process for results comparison purposes. This proposed work contributes towards finding an intelligent method of evaluating the transformer state as well as minimizing the number of tests without losing much information about the unit's actual condition. Therefore, this method facilitates deciding a wise course of action regarding the transformer: either maintain, repair, or replace.

Object detection utilizing frequency modulated continuous wave radar is becoming increasingly popular in the field of autonomous vehicles. Radar does not possess the same drawbacks seen by other emission-based sensors such as LiDAR, primarily the degradation or loss of return signals due to weather conditions such as rain or snow. Thus, there is a necessity for fully autonomous systems to utilize radar sensing applications in downstream decision-making tasks, generally handled by deep learning algorithms. Commonly, three transformations have been used to form range-azimuth-Doppler cubes in which deep learning algorithms could perform object detection. This method has drawbacks, specifically the pre-processing costs associated with performing multiple Fourier Transforms and normalization. We develop a network utilizing raw radar analog-to-digital converter output capable of operating in near real-time given the removal of all pre-processing. We obtain inference time estimates one-fifth of the traditional range-Doppler pipeline, decreasing from $\SI{156}{\milli\second}$ to $\SI{30}{\milli\second}$, and similar decreases in comparison to the full range-azimuth-Doppler cube. Moreover, we introduce hierarchical Swin Vision transformers to the field of radar object detection and show their capability to operate on inputs varying in pre-processing, along with different radar configurations, \textit{i.e.}, relatively low and high numbers of transmitters and receivers. Our network increases both average recall, and mean intersection over union performance by $\sim 6-7\%$, obtaining state-of-the-art F1 scores as a result on high-definition radar. On low-definition radar, we note an increase in mean average precision of $\sim 2.5\%$ over state-of-the-art range-Doppler networks when raw analog-to-digital converter data is used, and a $\sim5\%$ increase over networks using the full range-azimuth-Doppler cube.

In a conventional electronic ballast for a fluorescent lamp, inductor-capacitor-transformer tank circuit is used. A Piezoelectric Transformer (PT) can potentially be used to replace such a tank circuit to save space and cost. In the past, ballast design using a PT requires selecting a PT from available samples which are normally not matched to specific application and therefore resulting in poor performance. In this thesis, a design procedure was proposed for designing a PT tailored for a 120-V 32-W electronic ballast with high power factor, high efficiency and Zero-Voltage-Switching (ZVS) of the inverter transistors that drive the lamp. This involves selection of PT materials, determination of geometries and the number of physical layers of the PT. A radial mode piezoelectric transformer prototype based on this design process was fabricated by Face Electronics Inc. and was tested experimentally, the results showed that the ballast using this custom-made PT achieved high power factor, Zero-Voltage-Switching and a 83% overall efficiency. / Master of Science

Transients encountered by transformers in power stations during normal operation can have complex oscillatory overvoltages containing a large spectrum of frequency components. These transients can coincide with the natural frequencies of the transformers windings, leading to voltages that can be greater or more severe than the current factory proof tests. This may lead to insulation breakdown and catastrophic failures. Existing lumped parameter RLCG transformer models have been proven to be less accurate for very fast transient overvoltages (VFTO) with frequencies over 1 MHz. A white box model for transient analysis of transformer windings has been developed using Multiconductor Transmission Line (MTL) Theory. This model enables the simulation of natural frequencies of the transformer windings up to frequencies of several MHz, and can be used to compute voltages between turns by representing each turn as a separate transmission line. Both continuous and interleaved disk windings have been modelled and a comparison and validation of the results is presented. / May 2017

The multi-order response of the tuned secondary circuit of a Tesla transformer, following impulse excitation from its tuned primary circuit, is presented and analysed at the fundamental resonant frequency and at higher-order mode frequencies. A novel way of modifying the frequency response of the secondary coil is then investigated by utilising a technique normally applied to the design of a certain type of filter known as a helical filter. In general, these are used in radio and microwave frequency circuits in order to pass certain frequencies with little attenuation whilst significantly attenuating other frequencies. Design techniques, developed over several decades, modify and optimise the performance of such filters. The frequency response of the helical filter is modified by altering the geometry of the helical resonator component therein, which is typically in the form of an air-cored single-layer solenoid. A Tesla transformer whose secondary is constructed to be some form of single-layer solenoidal winding resonates at its designed frequency - its fundamental mode - but also at non-integer harmonics (higher-order anharmonic frequencies, also known as overtones). Those multi-order oscillatory voltages and currents energised in the secondary circuit have been identified and measured and research has determined the fundamental and higher-order mode frequencies and amplitudes for various experimental secondary winding configurations derived from helical filter design techniques. Applied to the Tesla transformer secondary winding, such techniques lead to a new design with a performance that is improved by the suppression of higher- order anharmonic frequencies whilst imparting little change to the fundamental response. It is anticipated that this feature will lead to Tesla transformers which exhibit enhanced spectral purity and which will be better suited to use in certain pulsed power applications than conventionally wound designs.

Even though there had been extensive research in Switching Mode Power Amplifier design their applications at industry level are quite limited. This is because a Fully-Integrated Switching Mode Power Amplifier using conventional active devices such as Bipolar Junction Transistors (BJT) or Metal Oxide Semiconductor Field Effect Transistor (MOSFET) is challenging due to the inherent design challenges in the Switching Power Amplifier design. A Fully-Integrated Differential Class-F2,3 Power Amplifier design is explored for this Thesis research. This Power Amplifier has a maximum theoretical efficiency of 90.7% but this value is reduced because of the switching nature of the active device, parasitic effects associated with layout and the quality factor of the passive components used. Waveform shaping required for a Class-F Power Amplifier is done using the stray inductances within a non-ideal transformer instead of individual inductors. This techniques effective reduces the foot prints of two inductors for the tuning network design and make a Fully-Integrated solution more practical.

Estudos de transitórios eletromagnéticos são importantes fontes de informação para que os transformadores sejam dimensionados de maneira correta. No entanto, para que tais estudos sejam bem sucedidos, os modelos utilizados devem refletir com fidelidade o comportamento do equipamento. Este trabalho mostra como os elementos do modelo de um transformador são influenciados pelas dimensões geométricas de sua parte ativa. Também introduz uma formulação alternativa, para o transformador saturável (STC) do ATP, desenvolvida dentro do programa MATLAB. Os ramos RL foram representados usando o Método de Integração Trapezoidal e a magnetização foi equacionada pelo Método da Compensação. Uma das contribuições que esta dissertação oferece é a possibilidade de identificar erros numéricos que ocorrem em simulações do ATP, bem como permitir a interpretação de resultados que apresentem oscilações numéricas. / Electromagnetic transient studies are an important source of information to develop transformer dimensioning. But, for the success of that purpose, it is important the models which are being used reflect with fidelity the behavior of the machine. This lecture presents how the transformer model elements are influenced by the active part geometrical dimensions. It also introduces an alternative formulation for the ATP saturable transformer (STC), written inside the MATLAB program. The RL branches are represented using the Trapezoidal Rule and the magnetization by the Compensation Method. One of the contributions of this dissertation is the possibility to identify numerical errors that occur in ATP simulations, and also permit numerical oscillatory results interpretation.

Uma série de equipamentos de eletrônica de potência é ligada à rede de corrente alternada através de transformadores, que exercem as funções de isolamento galvânico, elevação/abaixamento de tensão, etc. Estes transformadores podem sofrer saturação em condições transitórias (inrush) ou quando o conversor ligado a ele impõe valor médio não nulo de tensão. O problema é normalmente sanado nas malhas de controle do conversor, no entanto o ajuste destas malhas em ambiente de simulação computacional requer um modelo que represente adequadamente a característica de saturação do núcleo ferromagnético. Este trabalho apresenta um modelo de simulação computacional adequado para transformadores operando em saturação intensa, visando aplicações em projetos de equipamentos de eletrônica de potência com transformador monofásico alimentado por inversor. Além disso, é demonstrado um método experimental de obtenção da característica de saturação do núcleo, sem necessidade das suas características construtivas. Resultados de simulação obtidos são comparados com resultados experimentais para validação do modelo e método. / Several power electronics based equipments are connected to the alternating current network through transformers, which perform galvanic insulation, voltage increasing/lowering, etc. These transformers may experience saturation under transitory conditions (inrush), or when its converter imposes non-zero average voltage. Such problem is commonly treated in the converter\'s control loops, but the tuning of these loops in a computer simulation environment requires a transformer model that represents adequately the ferromagnetic core saturation effect. The current work proposes a computer simulation model that adequately describes the operation of the transformer at deep saturation, suitable for design of power electronics applications with single phase transformers connected to power inverter. Additionally, an experimental method for determining its core saturation characteristic is demonstrated. This method does not require previous knowledge of the core\'s constructive parameters. Simulation and experimental results are presented to confirm the validity of the model and the method.