Voltage boosting is required in many energy conversion applications, ranging from front-end solar energy system to Medium-Voltage DC (MVDC)-grid system, and from headlamp ballast of automobile to fuel cell powered system. This dissertation is dedicated to the development of voltage boosting technology. New circuits were found both in switched capacitor converters and inductor-based power converters. The former has the potential of full integration while the latter has combined the merits of conventional magnetic components and switched capacitor stages, leading to a new hybrid converter category. These contributions are captured in two major parts respectively: The first part discusses the switched capacitor converters and new discovery in voltage boosting technology as well as modeling technique. The second part introduces a number of new topologies and extensions in hybrid boosting converters.

Part I

Switched Capacitor Converter (SCC) is a special branch of power electronics converters which is composed of capacitors and switches without the participation of inductors/transformers. It potentially has lower electromagnetic interference (EMI), lighter weight, lower cost, higher energy density, and the promise for full integration.

In this part, a family of "Two-switch Boosting Switched-capacitor Converters (TBSC)" is introduced, which distinguishes itself from the prior arts by its symmetrical interleaved operation, reduced output ripple, low yet even voltage stress on components, and systematic expandability. Along with the topologies, a modeling technique based on charge-balance transient-calculation is formulated, which provokes the converter regulation method through duty cycle and frequency adjustment. The design guideline for high efficient TBSC is provided and regulation under high power condition is explored.

In addition, an enhanced accurate modeling technique considering the output capacitor effect is developed for simple two-phase SC converters.

Part II

Built upon the momentum of the "Two-switch Boosting Switched-capacitor Converters (TBSC)" circuit structure, another family of Hybrid Boosting Converters (HBC) is developed, featuring wide regulation range, symmetrical configuration, low component voltage stress, small output voltage ripple, and expandable structure. It integrates the inductive switching cores of various functionalities and control strategies with Bipolar Voltage Multiplier (BVM), resulting in the new breed of hybrid converters: Hybrid Boosting Converters (HBC). The proposed HBC family includes the basic HBC, symmetrical HBC, Isolated HBC, and tapped inductor HBC, all of which are analyzed in details and design considerations are provided.

In addition, the 3D DC-DC converter concept for high power application, inverter configuration and DC micro-grid based on HBC converters are developed.

The proposed HBCs are applicable in many areas such HID lamp driver, X-ray system, ion pumps, front-end photovoltaic energy system, and energy storage systems.

Part I: Resonant Switched-capacitor Converters with Wide Line Regulation Range

The traditional switched-capacitor converter (SCC) typically comes with the shortcomings of high transient current spike, hard-switched operation and limited line regulation range. In this thesis, a family of resonant two-switch boosting switched-capacitor converter (RTBSC) and a family of Ladder resonant switched-capacitor converter (RSC) with low transient current spike, soft-switching operation and wide line regulation range are proposed. By inserting a small resonant inductor, the additional benefit of reduction of one or two bulky capacitor banks to much smaller resonant capacitor(s) is obtained. By operating above the resonant frequency, the transistors are ZVS turned on and diodes are zero-current-switching (ZCS) turned off. Most importantly, the voltage-gain range is significantly expanded. This resonant configuration is applicable to TBSC, Ladder SCC and the Dickson SCC.

However, the regulation capability of the proposed RTBSC and Ladder RSC is still a challenge at light load condition. To ease this issue, a family of step-up resonant switched-capacitor converter (RSC) with continuously adjustable conversion ratio is proposed based on the Cuk Dual Resonance Core (DRC) concept. By adopting the on-time fixed frequency modulation, its conversion ratio can vary continuously and efficiently even at the light load condition (1/10 of the rated power). The configuration of Cuk DRC is applicable to the Series-parallel (SP) SCC, Fibonacci SCC, Voltage Doubler (VD) and Dickson SCC.

Part II: Control Maps for PWM plus Phase-shift Modulated DC-DC Converters

In this part, comprehensive analyses about the operation principle, power transmission capability, component stress and the zero-voltage-switching (ZVS) operation region for a three-level PWM plus phase-shift (PPS) modulated bidirectional push-pull converter and a family of two-level PPS modulated isolated bidirectional DC-DC converters (IBDCs) are presented. Based on the analyses, a set of control maps is established to assist converter operation and design in the low-circulating-current regions, high-power-transmission-capability regions, and ZVS operation regions. In the control maps, a circuit parameter is identified, providing a design limit to select switching frequency, leakage inductance, turns ratio and power level of the PPS modulated DC-DC converters.

Part I: Battery Management Converter System

Energy storage devices such as rechargeable batteries play a significant role in domestic, transportation, and industrial applications. The development of new battery technologies featuring high power and energy densities along with the strict safety, reliability, and lifetime requirements of the applications have brought the need for Battery Management Systems (BMSs). The first part of this thesis, is focused on a configuration called Battery Management Converter System (BMCS) that provides potential advantages over the conventional configurations such as lower cost and power loss and improved reliability and waveform quality. The balancing algorithm and grid connection control of such configuration for grid connected energy storage application has also been studied and developed in this part.

Part II: Multilevel Converter Topology and Control

Multilevel converters are the state-of-the-art power conversion systems for Medium Voltage (MV) high power applications. As MV drives for fans, pumps, mills, compressors, conveyors etc., multilevel converters have found their use in numerous industries such as mining, marine, metals, water, oil and gas, pulp and paper to name a few. As grid-connected inverters, multilevel converters are used to integrate renewable energy sources and energy storage systems into the grid and regulate the power flow and quality. In the second part of this thesis, the focus will be on the five-level converter topologies with a common dc-link as this feature suits them for many applications involving bidirectional flow of energy. New hybrid multilevel converter topologies along with control and modulation techniques are investigated and developed in this part. Improved trade off between cost and reliability and decreased converter loss are the advantages of the developed topologies and modulation techniques.

Power system reliability and security are crucial for the smooth functioning and sustainability of modern societies. The reliability of a power system refers to its ability to provide electricity on a continuous basis to meet the demands of consumers. Power system security refers to the ability to withstand sudden disturbances such as short circuits or unanticipated loss of elements. These aspects are of paramount importance to achieve economic stability and public safety, improve quality of life, and minimize environmental impact.

The first part of this dissertation focuses on the reliability and security of the utility side. It investigates the impact of active and reactive current injection on transient stability during the recovery from momentary cessation in power systems with high penetration of renewable energy resources. Theoretical analysis of the one-machine infinite-bus and multi-machine systems demonstrates the superiority of lagging current injection during the recovery period. A recovery strategy with an optimal transition from lagging to active current injection was recommended. The proposed strategy was tested by simulation on a one-machine infinite-bus system, a 9-bus power system, and a 39-bus power system.

The second part of the dissertation focuses on the reliability and security of the consumer side. It examines the postfault capability of a three-level neutral-point-clamped inverter under all possible fault scenarios. A proposed postfault operation method provides a comprehensive solution for operations after single and multiple-device failures, increasing the inverter reliability by 24%. Classification of postfault modulations uncovered previously unknown fault scenarios that can be addressed using the proposed control method. Two postfault modulations based on space vector modulation are proposed. The feasibility of the proposed control method is verified by simulation and an experiment with a 3.73 kW motor load.

Improvement on multilevel converters is fundamental to cover the increasing demands on power, efficiency, and quality. In this thesis, 3 recently developed 5-level converters are investigated with a focus on their thermal management and cooling design. Switching device temperature limits its maximum power and switching frequency for operation. This is one of the most important constraints, particularly, in multilevel converters, where the hottest device will limit the performance of the whole system. Then, the capability of the topology to uniformly distribute the power losses plays an important role.

Flying Capacitor-Active Neutral Point Clamped (FC-ANPC) [24, 25], Dual Flying Capacitor-Active Neutral Point Clamped (DFC-ANPC) [27, 29], and Mule [35] are simulated under different modulation indices and power factors. The resulting graphs can be used as a tool for the designer to know the worst thermal condition in a particular application. The study shows the ability of DFC-ANPC and Mule to naturally reach a lower temperature with its consequent advantages. Regarding DFC-ANPC and Mule, the differences are minimal and they depend on the situation. However, because of its higher number of states, the benefits of active control power losses balancing techniques could be slightly higher in Mule. Similar comparison has been done in relation to the fluctuation of temperature, which is an important cause of the device's failure due to thermal stress.

Finally, cooling systems strategies to cool multilevel converters are explained. It is mainly focused on commercially available cold plates, which has been analyzed pointing out the optimal flow rate.

Non-isolated high step-up and high step-down DC-DC conversions are highly demanded in many applications, including renewable energy systems, uninterrupted power supplies, data centers, high-end computers, etc. The high voltage-conversion-ratio requirement poses challenges on energy efficiency, power density, cost and reliability. This dissertation explores new high step-up and high step-down topologies for various applications.

First, a new family of interleaved high step-up converters integrating coupled-inductor and switched-capacitor are introduced for high current and ultra-high step-up applications. By integrating coupled-inductor and switched-capacitor techniques, the proposed converter achieves ultra-high step-up voltage gain without the need of extreme duty cycle or high turns ratio. Also, very low switch voltage stress can be achieved, thus low-voltage-rating MOSFETs with small on-resistance can be used to lower the conduction loss. Moreover, thanks to the interleaved operation at the input side, the input current is shared and low input current ripple is obtained. Furthermore, the coupled-inductor leakage energy can be recycled, which helps alleviate reverse recovery problem.

Second, multiphase interleaved high step-up converter with diode-capacitor voltage multiplier stages is presented, which is an extension of the previously reported two-phase current-fed Cockcroft–Walton multiplier. The multiphase configuration has flexible structure. It achieves high current handling capability, high voltage gain, low component voltage stress, low input current ripple and automatic phase current balancing. A three-phase converter in the multiphase family was analyzed and evaluated in detail.

Third, a single-phase coupled-inductor boost converter is presented for low current high boost applications. The voltage gain can be extended by the coupled inductor. Due to a independent inductor at the input, continuous input current is obtained, which significantly reduces the input capacitor, leading to increased power density and reliability. Also, the new converter has low component count, leading to low cost. Furthermore, recycled leakage energy and alleviated diode reverse recovery can be achieved.

Fourth, an improvement is proposed based on the proposed single-phase coupled-inductor boost converter. The improved converter not only retains the advantages of high voltage gain and continuous input current, but also achieves low semiconductor voltage stresses, zero-voltage switching and zero DC magnetizing current, which helps increase efficiency and power density.

Extensions of the above high step-up converters to high step-down operation are also discussed in the dissertation. Additionally, a 48V-to-1V high step-down converter is presented, where high step-down conversion is realized by an energy transfer capacitor and built-in transformer. Due to the interleaved technique, small values of output inductance can be used, leading to fast transient response. Furthermore, the use of a coupled inductor not only reduces the number of cores but also reduces inductor current ripples, leading to lower losses. Four switches are used in the converter and all can achieve zero-voltage-switching. Grounded gate drivers and boost strap drivers can be employed, avoiding the use of complex isolated gate drivers. In addition, if one of the switches fails, the high input voltage is blocked from the output to keep the load safe.

Part I: Dynamic Voltage Restorer

In the present power grids, voltage sags are recognized as a serious threat and a frequently occurring power-quality problem and have costly consequence such as sensitive loads tripping and production loss. Consequently, the demand for high power quality and voltage stability becomes a pressing issue. Dynamic voltage restorer (DVR), as a custom power device, is more effective and direct solutions for "restoring" the quality of voltage at its load-side terminals when the quality of voltage at its source-side terminals is disturbed.

In the first part of this thesis, a DVR configuration with no need of bulky dc link capacitor or energy storage is proposed. This fact causes to reduce the size of the DVR and increase the reliability of the circuit. In addition, the proposed DVR topology is based on high-frequency isolation transformer resulting in the size reduction of transformer. The proposed DVR circuit, which is suitable for both low- and medium-voltage applications, is based on dc-ac converters connected in series to split the main dc link between the inputs of dc-ac converters. This feature makes it possible to use modular dc-ac converters and utilize low-voltage components in these converters whenever it is required to use DVR in medium-voltage application. The proposed configuration is tested under different conditions of load power factor and grid voltage harmonic. It has been shown that proposed DVR can compensate the voltage sag effectively and protect the sensitive loads.

Following the proposition of the DVR topology, a fundamental voltage amplitude detection method which is applicable in both single/three-phase systems for DVR applications is proposed. The advantages of proposed method include application in distorted power grid with no need of any low-pass filter, precise and reliable detection, simple computation and implementation without using a phased locked loop and lookup table. The proposed method has been verified by simulation and experimental tests under various conditions considering all possible cases such as different amounts of voltage sag depth (VSD), different amounts of point-on-wave (POW) at which voltage sag occurs, harmonic distortion, line frequency variation, and phase jump (PJ). Furthermore, the ripple amount of fundamental voltage amplitude calculated by the proposed method and its error is analyzed considering the line frequency variation together with harmonic distortion. The best and worst detection time of proposed method were measured 1ms and 8.8ms, respectively. Finally, the proposed method has been compared with other voltage sag detection methods available in literature.

Part 2: Power System Modeling for Renewable Energy Integration

As power distribution systems are evolving into more complex networks, electrical engineers have to rely on software tools to perform circuit analysis. There are dozens of powerful software tools available in the market to perform the power system studies. Although their main functions are similar, there are differences in features and formatting structures to suit specific applications. This creates challenges for transferring power system circuit models data (PSCMD) between different software and rebuilding the same circuit in the second software environment. The objective of this part of thesis is to develop a Unified Platform (UP) to facilitate transferring PSCMD among different software packages and relieve the challenges of the circuit model conversion process. UP uses a commonly available spreadsheet file with a defined format, for any home software to write data to and for any destination software to read data from, via a script-based application called PSCMD transfer application. The main considerations in developing the UP are to minimize manual intervention and import a one-line diagram into the destination software or export it from the source software, with all details to allow load flow, short circuit and other analyses. In this study, ETAP, OpenDSS, and GridLab-D are considered, and PSCMD transfer applications written in MATLAB have been developed for each of these to read the circuit model data provided in the UP spreadsheet. In order to test the developed PSCMD transfer applications, circuit model data of a test circuit and a power distribution circuit from Southern California Edison (SCE) - a utility company - both built in CYME, were exported into the spreadsheet file according to the UP format. Thereafter, circuit model data were imported successfully from the spreadsheet files into above mentioned software using the PSCMD transfer applications developed for each software.

After the SCE studied circuit is transferred into OpenDSS software using the proposed UP scheme and developed application, it has been studied to investigate the impacts of large-scale solar energy penetration. The main challenge of solar energy integration into power grid is its intermittency (i.e., discontinuity of output power) nature due to cloud shading of photovoltaic panels which depends on weather conditions. In order to conduct this study, OpenDSS time-series simulation feature, which is required due to intermittency of solar energy, is utilized. In this study, the impacts of intermittency of solar energy penetration, especially high-variability points, on voltage fluctuation and operation of capacitor bank and voltage regulator is provided. In addition, the necessity to interpolate and resample unequally spaced time-series measurement data and convert them to equally spaced time-series data as well as the effect of resampling time-interval on the amount of error is discussed. Two applications are developed in Matlab to do interpolation and resampling as well as to calculate the amount of error for different resampling time-intervals to figure out the suitable resampling time-interval. Furthermore, an approach based on cumulative distribution, regarding the length for lines/cables types and the power rating for loads, is presented to prioritize which loads, lines and cables the meters should be installed at to have the most effect on model validation.

Derived from classic DC-AC multilevel converters, hybrid multilevel converters can combine distinct advantages, thus receiving increasing attention for their widespread industrial applications recently. The Active Neutral Point Clamped (ANPC) converter offers redundant switching states to achieve a better control and more even loss distribution compared to Neutral Point Clamped (NPC) converter. However, in terms of five-level converters, currently the only topology applied in industrial application is the Flying-Capacitor based Neutral Point Clamped converter (FCANPC), which gives a relatively good performance with a simple structure. Despite its advantages, the uneven loss distribution cannot be disregarded as a substantial drawback to affect the power processing capability. The Dual Flying Capacitor Neutral Point Clamped (DFCANPC) converter has been newly proposed to provide better loss balancing and "soft commutation" to have a higher power capability and efficiency.

In this thesis, traditional multilevel topologies such as Cascaded H-Bridge converter, Diode Clamped converter, and Flying Capacitor converter are summarized, followed by popular modulation schemes including carrier-based modulation, Space Vector Modulation and Selective Harmonic Elimination modulation. Thermal models and cooling strategies are then briefly discussed. As the main focus, thermal analysis and comparison regarding the five-level FCANPC converter and the newly proposed DFCANPC converter is reported in this thesis. A PSIM model has been built for verification of loss balancing performance of the DFCANPC converter.

This thesis applies a fixed ON-time switching technique to a resonant switched-capacitor converter, called the 3X RTBSC-A converter, which is studied for maximum power point tracking (MPPT) applications. The previous converter, the 3X RTBSC converter, was unable to achieve full-range voltage gain regulation for relatively light loads. However, with a new switching technique known as Switching Technique A, full-range voltage gain regulation for all loads is completely possible. In addition, the converter achieves zero-voltage switching (ZVS) to retain a high efficiency. A mathematical analysis of the ideal circuit was completed to show the conducting states, voltage gain derivation, and component stresses. Under this analysis, the prototype circuit was designed with a maximum power level of 162 W and a resonant frequency of 100 kHz. Simulations in the LTspice software proved that the circuit would not be able to achieve full-range voltage gain regulation with regular 50% duty cycle switching. In addition, details are provided in the design of the prototype's control circuit and driver circuit used to create and execute Switching Technique A. An open loop experiment demonstrates that the voltage gain of the prototype circuit closely resembles the waveforms calculated from the ideal gain formula. Also, a relatively high efficiency is recorded for each load tested with a maximum recorded efficiency of 0.9544. In addition, examination of the steady state waveforms prove that the prototype circuit is executing ZVS. To prepare the prototype for the MPPT experiment, a sensing circuit is designed to feed data to the microcontroller. A simple perturb and observe (P&O) algorithm is programmed to find the maximum power point (MPP). The specifications of the solar panel are also presented. The results of the MPPT experiment demonstrate that the MPP found by the MPPT algorithm closely follows the actual MPP that was found manually. In addition, the prototype achieved efficiencies between 0.87 and 0.92 during the MPPT experiment. The 3X RTBSC-A converter still has the potential to achieve higher efficiencies as indicated by the report on the original circuit.