Presently, wireless sensor nodes are widely used and the lifetime of the system is becoming the biggest problem with using this technology. As more and more low power products have been used in WSN, energy harvesting technologies, based on their own characteristics, attract more and more attention in this area. But in order to design high energy efficiency, low cost and nearly perpetual lifetime micro energy harvesting system is still challenging. This thesis proposes a new way, by applying three factors of the system, which are the energy generation, the energy consumption and the power management strategy, into a theoretical model, to optimally design a highly efficient micro energy harvesting system in a real environment. In order to achieve this goal, three aspects of contributions, which are theoretically analysis an energy harvesting system, practically enhancing the system efficiency, and real system implementation, have been made. For the theoretically analysis, the generic architecture and the system design procedure have been proposed to guide system design. Based on the proposed system architecture, the theoretical analytical models of solar and thermal energy harvesting systems have been developed to evaluate the performance of the system before it being designed and implemented. Based on the model's findings, two approaches (MPPT based power conversion circuit and the power management subsystem) have been considered to practically increase the system efficiency. As this research has been funded by the two public projects, two energy harvesting systems (solar and thermal) powered wireless sensor nodes have been developed and implemented in the real environments based on the proposed work, although other energy sources are given passing treatment. The experimental results show that the two systems have been efficiently designed with the optimization of the system parameters by using the simulation model. The further experimental results, tested in the real environments, show that both systems can have nearly perpetual lifetime with high energy efficiency.

Simulations are a good method for the verification of the correct operation of solar-powered sensor nodes over the desired lifetime. They do, however, require accurate models to capture the influences of the loads and solar energy harvesting system. Artificial neural networks promise a simplification and acceleration of the modeling process in comparison to state-of-the-art modeling methods. This work focuses on the influence of the modeling process's different configurations on the accuracy of the model. It was found that certain parameters, such as the network's number of neurons and layers, heavily influence the outcome, and that these factors need to be determined individually for each modeled harvesting system. But having found a good configuration for the neural network, the model can predict the supercapacitor's charge depending on the solar current fairly accurately. This is also true in comparison to the reference models in this work. Nonetheless, the results also show a crucial need for improvements regarding the acquisition and composition of the neural network's training set.

Jellyfish have inhabited this planet for millions of years and are the oldest known metazoans that swim using muscles. They are found in freshwater sources and in oceans all over the world. Over millions of years of evolution, they have adapted to survive in a given environment. They are considered as one of the most energy efficient swimmers. Owing to these characteristics, jellyfish has attracted a lot of attention for developing energy efficient unmanned undersea vehicles (UUVs). The goal of this thesis is to provide understanding of the different physical mechanisms that jellyfish employs to achieve efficient swimming by using analytical and computational models. The models were validated by using the experimental data from literature. Based upon these models refinements and changes to engineering vehicles was proposed that could lead to significant enhancement in propulsion efficiency. In addition to the propulsion, the thesis addresses the practical aspects of deploying a jellyfish-inspired robotic vehicle by providing insights into buoyancy control and energy generation. The thesis is structured in a manner such that propulsive and structural models inspired from the natural animal were systematically combined with the practical aspects related to ionic diffusion driven buoyancy control system and thermal -- magnetic energy harvesting system. Jellyfish morphology, swimming mechanism and muscle architecture were critically reviewed to accurately describe the natural behavior and material properties. We provide full understanding of mesoglea, which plays most significant role towards swimming performance, in terms of composition, mechanical properties and nonlinear dynamics. Different jellyfish species exhibit different microstructure of mesoglea and thus there is a wide variety of soft materials. Mechanical properties of collagen fibers that form the main constituent toward imparting elasticity to mesoglea were reviewed and analyzed. The thesis discusses the theoretical models describing the role of structure of mesoglea towards its mechanical properties and explains the variation occurring in stiffness under given experimental environment. Muscle architecture found in jellyfish, nerve nets and its interconnection with the muscles were investigated to develop comprehensive understanding of jellyfish propulsion and its reaction to external stimuli. Different muscle arrangements were studied including radial, coronal muscle, and coronal-muscles-with-breaks in-between them as observed in Cyanea capillata. We modeled these muscle arrangements through finite element modeling (FEM) to determine their deformation and stroke characteristics and their overall role in bell contraction. We found that location and arrangement of coronal muscle rings plays an important role in determining their efficient utilization. Once the understanding of natural jellyfish was achieved, we translated the findings onto artificial jellyfish vehicle designed using Bio-inspired Shape Memory Alloy Composite (BISMAC) actuators. Detailed structural modeling was conducted to demonstrate deformation similar to that of jellyfish bell. FEM model incorporated hyperelastic behavior of artificial mesoglea (Ecoflex-0010 RTV, room temperature vulcanizing silicone with shore hardness (0010)), experimentally measured SMA temperature transformation, gravity and buoyancy forces. The model uses the actual control cycle that was optimized for driving the artificial jellyfish vehicle "robojelly". Using a comparative analysis approach, fundamental understanding of the jellyfish bell deformation, thrust generation, and mechanical efficiency were provided. Meeting energy needs of artificial vehicle is of prime importance for the UUVs. Some jellyfish species are known to use photosynthesis process indirectly by growing algae on their exumbrella and thereby utilizing the sunlight to generate energy. Inspired by this concept, an extensive model was developed for harvesting solar energy in underwater environment from the jellyfish bell structure. Three different species were modeled for solar energy harvesting, namely A.aurita, C.capillata and Mastigia sp., using the amorphous silicon solar cell and taking into account effect of fineness ratio, bell diameter, turbidity, depth in water and incidence angle. The models shows that in shallow water with low turbidity a large diameter vehicle may actually generate enough energy as required for meeting the demand of low duty cycle propulsion. In future, when the solar energy harvesting technology based upon artificial photosynthesis, referred to as "dye-sensitized solar cells", matures the model presented here can be easily extended to determine its performance in underwater conditions. In order to supplement the energy demand, a novel concept of thermal -- magnetic energy harvesting was developed and extensively modeled. The proposed harvester design allows capturing of even small temperature differences which are difficult for the thermoelectrics.  A systematic step-by-step model of thermo-magnetic energy harvester was presented and validated against the experimental data available in literature. The multi-physics model incorporates heat transfer, magnetostatic forces, mechanical vibrations, interface contact behavior, and piezoelectric based energy converter. We estimated natural frequency of the harvester, operating temperature regimes, and electromechanical efficiency as a function of dimensional and physical variables. The model provided limit cycle operation regimes which can be tuned using physical variables to meet the specific environment. Buoyancy control is used in aquatic animals in order to maintain their vertical trajectory and travel in water column with minimum energy expense. Some crustaceans employ selective ion replacement of heavy or lighter ions in their dorsal carapace. A model of a buoyancy chamber was developed to achieve similar buoyancy control using electro-osmosis. The model captures all the essential ionic transport and electrochemistry to provide practical operating cycle for the buoyancy engine in the ocean environment. / Ph. D.

Microbial electrosynthesis (MES) has recently emerged as a novel biotechnology platform for value-added product generation from waste CO2 stream. Integrating MES technology with renewable energy sources for both CO2 valorization and renewable energy storage is regarded as one type of artificial photosynthesis and a perfect example of circular bioeconomy. However, several challenges remain to be addressed to scale-up MES as a feasible process for chemical production, which include enhanced production rate, reduced energy consumption and excellent resistance to external fluctuations. To fill these knowledge gaps, different in-depth approaches were proposed in this dissertation by optimizing the cathode architecture, CO2 flow rates and utilizing efficient photoelectrode to improve MES performance and stability. A novel cathode design, made of conductive hollow fiber membrane, was developed in this dissertation to improve CO2 availability at MES cathode surface via direct CO2 delivery to chemolithoautotrophs through the pores in the hollow fibers. By modifying the hollow fiber surface with carbon nanotubes (CNTs), higher bioproduct formation was achieved with excellent faradaic efficiencies, which could be attributed to the improved surface area for bacterial adhesion and the reduction of cathodic electron transfer resistance. Since CO2 flow rate from industrial facilities typically varies over time, this hollow-fiber architecture was also applied to test the resistance of MES systems to CO2 flow rate fluctuation. Stepwise increase of CO2 flow rates from 0.3 ml/min to 10 ml/min was tested and the effect of CO2 flow rate fluctuations was evaluated in terms of biochemical generation and microbial community. MES was further integrated with renewable energy supply for both energy storage and CO2 transformation into biofuels and biochemicals. Stable MES photoanode, based on molybdenum-doped bismuth vanadate deposited on fluorine-doped tin oxide glass (FTO/BiVO4/Mo), was prepared for efficient solar energy harvesting and overpotential reduction for oxygen evolution reaction (OER), which contributed to one of the highest solar-to-biochemical conversion efficiencies ever reported for photo-assisted MES systems. The applied nature of this dissertation with fundamental insights is of great importance to bring MES one step closer to full-scale applications and enable MES technology to be economically more viable for renewable energy storage and CO2 valorization.

A wireless imaging system for monitoring strawberry fields provides enough quality image data for computer vision algorithms to make meaningful yield predictions. This report contains a design for a wireless sensor network modified with mesh networking techniques to extend coverage range and a solar energy harvesting system to improve sensor node lifetime. A two hop system with six nodes is implemented in a laboratory environment validating the communication systems integrity over an 800’ range. Moving from a primary battery system to solar energy harvesting increases the module lifetime indefinitely.

The proposed energy harvesting circuit system is designed for indoor solar environment especially for factories where the light energy is abundant and stable. The designed circuits are intended to power wireless sensor nodes (WSNs) or other computing unit such as microcontrollers or DSPs to provide a power solution for Internet of Things (IoTs). The proposed circuit can extract maximum power from the PV panel by utilizing the maximum power point tracking (MPPT) technique. The power stage is a synchronous dual-input dual-output non-inverting buck-boost converter operating in discontinuous conduction mode (DCM) and constant on-time pulse skipping modulation (COT-PSM) to achieve voltage regulation and maximum power delivery to the load. Battery is used as secondary input also as secondary output to achieve a longer lifecycle, a fast load response time and support higher load conditions. The proposed MPPT technique doesn't require any current sensor or computing units. Fully digitalized simple circuits are used to achieve sampling, store, and comparing tasks to save power. The whole circuits including power stage and control circuits are designed and will fabricate in TSMC BCDMOS 180 nm process. The circuits are verified through schematic level simulations and post-layout simulations. The results are validated to prove the proposed circuit and control scheme work in a manner. / Master of Science / With the growing energy demands, the efficient energy conversion systems caught great attentions. Especially, in the era of Internet of Things, powering those wireless devices can be extremely difficult. Nowadays, lots of devices such as consumer electronics, wireless sensor nodes, computing and mission system etc. are still powered by the batteries. Regular changing the batteries of those devices can be inconvenient or expensive. Energy harvesting provides a good solution to this issue because there are lots of ambient energy source is available. To design an energy efficient energy harvesting circuit system can help extend the device lifecycle per charging cycle. Even with some specific energy source which power scale is high enough, meanwhile the load doesn’t require too much power, the devices can be power-independent or standalone. In this work, the proposed circuit targets for indoor solar energy harvesting via solar panel. The target powering devices are wireless sensor nodes (WSNs). Meanwhile, WSNs can monitor the temperature, humidity, pressure, noise level etc. The proposed circuit design combines the power stage and control circuit on an integrated chip (IC), only few components are off-chip. It provides a very compact, endurable, and economical solution to the current IoT powering issue.

Structural health monitoring systems provide a promising route to real-time data for analyzing the current state of large structures. In the wake of two high-profile bridge collapses due to an aging highway infrastructure, the interest in implementing such systems into fracture-critical and structurally deficient bridges is greater now than at any point in history. Traditionally, these technologies have not been cost-effective as bridges lack existing wiring architecture and the addition of this is cost prohibitive. Modern wireless sensor networks (WSN) now present a viable alternative to traditional networking; however, these systems must incorporate localized power sources capable of decade-long operation with minimal maintenance. To this end, this thesis explores the development of two energy harvesting systems capable of long-term bridge deployment with minimal maintenance. First, an electromagnetic, linear, vibrational energy harvester is explored that utilizes the excitations from passing traffic to induce motion in a translating permanent magnet mass. This motion is then converted to electrical energy using Faraday’s law of induction. This thesis presents a review of vibrational energy harvesting literature before detailing the process of designing, simulating, prototyping, and testing a selected design. Included is an analysis of the effects of frequency, excitation amplitude, load, and damping on the power production potential of the harvester. Second, a solar energy harvester using photovoltaic (PV) panels is explored for powering the critical gateway component of the WSN responsible for data aggregation. As solar energy harvesting is a more mature technology, this thesis focuses on the methodologies for properly sizing a solar harvesting system and experimentally validating the selected design. Fabrication of the prototype system was completed and field testing was performed in Austin, TX. The results validate the selected system’s ability to power the necessary 14 W DC load with a 0° panel azimuth angle (facing direct south) and 45° tilt. / text

<div>My dissertation research is in the field of plasmonics. Specifically, my focus is on the use of plasmonics for various applications such as solar energy harvesting and optically addressable magnetic memory devices. Plasmonics is the study of collective oscillations of free electrons in a metal coupled to an electromagnetic field. Such oscillations are characterized by large electromagnetic field intensities confined in nanoscale volumes and are called plasmons. Plasmons can be excited on a thin metal film, in which case they are called surface plasmon polaritons or in nanoscale metallic particles, in which case they are called localized surface plasmon resonances. Researchers have taken advantage of this electromagnetic field enhancement resulting from the excitation of plasmons in metallic structures and demonstrated phenomenon such as plasmon-assisted photocatalysis, plasmon-induced local heating, plasmon-enhanced chemical sensing, optical modulators, nanolasers, etc.</div><div>In the first half of my dissertation, I study the role of plasmonics in hydrogen production from water using solar energy. Hydrogen is believed to be a very viable source of alternative green fuel to meet the growing energy demands of the world. There are significant efforts in government and private sectors worldwide to implement hydrogen fuel cells as the future of the automotive and transportation industry. In this regard, water splitting using solar energy to produce hydrogen is a widely researched topic. It is believed that a Solar-to-Hydrogen (STH) conversion efficiency of 10% is good enough to be considered for practical applications. Iron oxide (alpha-Fe2O3) or hematite is one of the candidate materials for hydrogen generation by water splitting with a theoretical STH efficiency of about 15%. In this work, I experimentally show that through metallic gold nanostructures we can enhance the water oxidation photocurrent in hematite by two times for above bandgap wavelengths, thereby increasing hydrogen production. Moreover, I also show that gold nanostructures can result in a hematite photocurrent enhancement of six times for below bandgap wavelengths. The latter, I believe, is due to the excitation of plasmons in the gold nanostructures and their subsequent decay into hot holes which are harvested by hematite.</div><div>The second part of my dissertation involves data storage in magnetic media. Memory devices based on magnetic media have been widely investigated as a compact information storage platform with bit densities exceeding 1Tb/in2. As the size of nanomagnets continue to reduce to achieve higher bit densities, the magnetic fields required to write information in these bits increases. To counter this, the field of heat-assisted magnetic recording (HAMR) was developed where a laser is used to locally heat up a magnet and make it susceptible to smaller magnetic switching fields. About two decades ago, it was realized that a single femtosecond laser pulse can switch magnetic media and therefore could be used to write information in magnetic bits. This field is now known as All-Optical Magnetic Switching (AOMS). My research aims to bring together the two fields of HAMR and AOMS to create optically addressable nanomagnets for information storage. Specifically, I want to show that plasmonic resonators can couple the laser field to nanomagnets more efficiently. This can therefore be used not only to heat the nanomagnets but also switch them with lower optical energy compared to free-standing nanomagnets without any plasmonic resonator. The results of my research show that by coupling metallic resonators, supporting surface plasmons, to nanomagnets, one can reduce the light intensity required for laser induced magnetization reversal.</div>