The DC House project at Cal Poly State University faces a challenge of supplying DC voltage to household appliances. Each appliance in the DC House constitutes a DC load that has a unique voltage and power rating, hence the need to develop a smart DC wall plug that will automatically adjust to the operating voltage required by any DC load. This thesis entails a proof of concept design of the smart DC wall plug which can automatically detect an appliance’s voltage rating. The design employs a dc-dc converter in conjunction with a microcontroller to sense load current to properly adjust the required load voltage. Hardware implementation to demonstrate the functionality of the smart wall plug was developed. Results performed on several dc loads show that the smart wall plug is able to adjust to the required load voltage within an acceptable range. An algorithm to improve the accuracy was attempted and presented along with the results. Further recommendations to improve the current design will also be discussed.

Load transients are prevalent in every electronic device including semiconductor memory, card readers, microprocessors, disc drives, piezoelectric devices, and digitally based systems. They are capable of producing voltage stress, introducing noise, and degrading device functionality. In order to avoid damage to the device, a feedback control loop is implemented with system compensation to regulate the output voltage deviations by the converter. Because designing compensation networks can be rather complicated, DC-DC converters with integrated feedback control topologies help minimize design time and complexity of converter compensation at the expense of design flexibility. This thesis widens the limitations of an integrated DC-DC converter with a stability optimization technique that utilizes the feedback network to create a phase boost centered at the bandwidth of the converter to increase the phase margin and improve its transient response. Ideal modeling verifies stability optimization while non-ideal modeling that introduces PCB parasitics to the control loop suggest an additional phase boost in the feedback network. Experimental data confirms this non-deal model for parasitic capacitances higher than calculated. The modified non-ideal model shows more accuracy compared to the experimental data which indicates that there may be PCB parasitics that is unaccounted for. Modeling the modified non-ideal model to high orders may yield more accuracy. This thesis gives both DC-DC converter and PCB layout designers insight and considerations into PCB effects on the stability of DC-DC converters and the optimization of integrated compensation.

This report demonstrates improvements made in battery charging and battery management technology through the design of a universal programmable battery charger with optional battery management system attachment. This charger offers improvements in charge efficiency and unique battery charging algorithms to charge a variety of battery chemistries with variety of power requirements. Improvements in efficiency result from a synchronous Buck Controller topology as compared to previous universal chargers that use asynchronous Buck-Boost Converter topologies. This battery charger also surpasses current universal battery chargers by offering different charge modes for different battery chemistries. Charge modes provide the user an option between extending the life of the battery by selecting a mode with a slower, less stressful charge rate or a shorter charge time with a fast, more stressful charging mode. The user can also choose a charge mode in which the battery charges to full capacity, resulting in maximum runtime or a less than full capacity, which puts less stress on the battery thus extending the lifetime. Ultimately, this system permits weighing the performance tradeoff of battery lifetime and charge time. The optional battery management system attachment offers more precise monitoring of each cell and cell balancing for Li-Ion batteries. This further enhances the performance of the charger when integrated, but is not necessary for charger operation. The battery charger consists of three subcircuits: A microcontroller unit, a power stage, and a current sensing circuit. A C2000 Piccolo F28069 microcontroller controls a LM5117 Buck Controller by injecting a pulse-width modulated signal into the feedback node controlling the output of the buck to set a constant current or constant voltage thus creating a programmable battery charger. The pulse-width modulated signal changes according to charge algorithms created in software for specific battery chemistries and charge requirements. An analog-to-digital converter on the microcontroller monitors battery voltage by using a voltage divider and an INA169 current shunt monitor, which outputs a voltage corresponding to the charge current to another analog-to-digital converter on the microcontroller, monitors the charge current. This allows the charger program to maintain correct and safe charging conditions for each charge mode in addition to measuring output power. Lights on the microcontroller display a real-time status to the user of which portion of the charge profile the charger is in. A solid red light means the charger is in the constant current portion of the charge profile. A blinking red light means the charger is in the constant voltage portion. No red light means the battery charger finished and the battery is currently charged above nominal voltage. The battery charger works with the battery management system in the next section to provide ultimate battery charging and managing capabilities. The battery management system consists of two subcircuits: A microcontroller and a battery monitoring circuit. The MSP430FR5969 microcontroller unit communicates with BQ76PL536 battery management integrated circuits to create a battery management system that monitors data such as cell voltage, pack voltage, pack temperature, state of charge, fault statuses, alert statuses, and a variety of other useful cell parameters. This data displays on a liquid crystal display screen through different menu options. The user scrolls through the menus using a capacitive touch slider on the microcontroller unit and selects a given option using the option select button. A cell balance mode allows the user to check the balance of the cells and allows cell balancing if the cells differ by more than a set threshold.

Every year the automotive industry strives to increase fuel efficiency in vehicles. When most vehicles are designed, fuel efficiency cannot always come first. The Shell Eco-marathon changes that by challenging students everywhere to develop the most fuel-efficient vehicle possible. There are many different factors that affect fuel efficiency, and different teams focus on different vehicle parameters. Currently, there is no straightforward design tool that can be used to help in Shell Eco-marathon vehicle design. For this reason, it is difficult to optimize every vehicle parameter for maximum fuel efficiency. In this study, a simulation is developed by using basic vehicle models and experimental data to accurately represent any prototype-class vehicle in the Shell Eco-marathon. This simulation is verified using different experimental data from an on-vehicle data acquisition system. An easy-to-use design tool is developed, and this tool is used to optimize driving strategy and final drive ratio to maximize fuel efficiency.

The SINDy (Sparse Identification of Non-linear Dynamics) algorithm is a method of turning a set of data representing non-linear dynamics into a much smaller set of equations comprised of non-linear functions summed together. This provides a human readable system model the represents the dynamic system analyzed. The SINDy algorithm is important for a variety of applications, including high precision industrial and robotic applications. A Hardware Accelerator was designed to decrease the time spent doing calculations. This thesis proposes an efficient hardware accelerator approach for a broad range of applications that use SINDy and similar system identification algorithms. The accelerator is leverages both systolic arrays for integrated neural network models with other numerical solvers. The novel and efficient reuse of similar processing elements allows this approach to only use a minimal footprint, so that it could be added to microcontroller devices or implemented on lower cost FPGA devices. Our proposed approach also allows the designer to offload calculations onto edge devices from controller nodes and requires less communication from those edge devices to the controller due to the reduced equation space.

Current consumption measurements are useful in a wide variety of applications, including power monitoring and fault detection within a lithium battery management system (BMS). This measurement is typically taken using either a shunt resistor or a Hall-effect current transducer. Although both methods have achieved accurate current measurements, shunt resistors have inherent power loss and often require isolation circuitry, and Hall-effect sensors are generally expensive. This work explores a novel alternative to sensing battery current by measuring terminal voltages and cell temperatures and using an unknown input observer (UIO) to estimate the battery current. An accurate model of a LiFePO4 cell is created and is then used to characterize a model of the proposed current estimation technique. Finally, the current estimation technique is implemented in hardware and tested in an online BMS environment. Results show that the current estimation technique is sufficiently accurate for a variety of applications including fault detection and power profiling.

A successful power system in military applications (warship, aircraft, armored vehicle etc.) must operate acceptably under a wide range of conditions involving different loading configurations; it must maintain war fighting ability and recover quickly and stably after being damaged. The introduction of energy storage for the power system of an electric warship integrated engineering plant (IEP) may increase the availability and survivability of the electrical power under these conditions. Herein, the problem of energy storage control is addressed in terms of maximizing the average performance. A notional medium-voltage dc system is used as the system model in the study. A linear programming model is used to simulate the power system, and two sets of states, mission states and damage states, are formulated to simulate the stochastic scenarios with which the IEP may be confronted. A genetic algorithm is applied to the design of IEP to find optimized energy storage control parameters. By using this algorithm, the maximum average performance of power system is found.

Scanning-electron-beam lithography (SEBL) is the primary technology to generate arbitrary features at the nano-scale. However, pattern placement accuracy still remains poor compared to its resolution due to the open-loop nature of SEBL systems. Vibration, stray electromagnetic fields, deflection distortion and hysteresis, substrate charging, and other factors prevent the electron-beam from reaching its target position and one has no way to determine the actual beam position during patterning with conventional systems. To improve the pattern placement accuracy, spatial-phase-locked electron-beam lithography (SPLEBL) provides feedback control of electron-beam position by monitoring the secondary electron signal from electron-transparent fiducial grids on the substrate. While scanning the electron beam over the fiducial grids, the phase of the grid signal is analyzed to estimate the electron-beam position error; then the estimates are sent back to beam deflection system to correct the position error. In this way, closed-loop control is provided to ensure pattern placement accuracy. The implementation of spatial-phase-locking on high speed field-programmable gate array (FPGA) provides a low-cost method to create a nano-manufacturing platform with 1 nm precision and significantly improved throughput. Shot-to-shot, or pixel-to-pixel, dose variation during EBL is a significant practical and fundamental problem. Dose variations associated with charging, electron source instability, optical system drift, and ultimately shot noise in the beam itself conspire to increase critical dimension variability and line width roughness and to limit the throughput. It would be an important improvement to e-beam patterning technology if real-time feedback control of electron-dose were provided to improve pattern quality and throughput even beyond the shot noise limit. A novel approach is proposed in this document to achieve the real-time dose control based on the measurement of electron arrival at the sample to be patterned, rather than from the source or another point in the electron-optical system. A dose control algorithm, implementation on FPGA, and initial experiment results for the real-time feedback dose control on the e-beam patterning tool is also presented.

The resilience of manufacturing enterprises is an important research topic, since disruptions have severe effects on the normal operation of manufacturing enterprises, especially as manufacturing supply chains become global. Although many case studies have been carried out to address resilience in organizations, a systematic method to model and analyze the resilience dynamics in manufacturing enterprises is not well developed. This study is intended to conduct research on quantitative analysis and control for resilience. After reviewing the literature addressing resilience, a modeling framework is presented to characterize the resilience of a manufacturing enterprise responding to disruptive events, which includes inventory ow between enterprise nodes, different costs, resource, demand, etc. Each node within the network is represented as a dynamic model with associated costs of production and inventory. This mathematical model is the foundation of quantitative analysis and control. With this model, an optimal control problem is formulated, by which the control can be solved to achieve minimum cost. Several different types of systems are defined and analyzed in this work. We develop the approach of aggregation to simplify the network structures. The study is mainly focused on two categories of network systems: serial network systems and assembly tree network systems. The analysis on these two categories covers two conditions: in discrete time domain without considering capacities, and in continuous time domain with considering capacities. The methods to determining optimal operations are developed under different conditions. In the serial network systems analysis, a practical case study is introduced to show the corresponding method developed. Finally, the problems are discussed for future research. Based on the results of these analyses, we present optimal control policies for resilience. Our method can support the analysis of the impact of disruptions, and the development of control strategies that reduce the impact of the disruption.

Energy consumption has increased drastically during the last century. Currently, the worldwide energy consumption is about 17.4 TW and is predicted to reach 25 TW by 2035. Solar energy has emerged as one of the potential renewable energy sources. Since its first physical recognition in 1887 by Adams and Day till nowadays, research in solar energy is continuously developing. This has lead to many achievements and milestones that introduced it as one of the most reliable and sustainable energy sources. Recently, the International Energy Agency declared that solar energy is predicted to be one of the major electricity production energy sources by 2035. Enhancing the efficiency and lifecycle of photovoltaic (PV) modules leads to significant cost reduction. Reducing the temperature of the PV module improves its efficiency and enhances its lifecycle. To better understand the PV module performance, it is important to study the interaction between the output power and the temperature. A model that is capable of predicting the PV module temperature and its effects on the output power considering the individual contribution of the solar spectrum wavelengths significantly advances the PV module designs toward higher efficiency. In this work, a thermoelectrical model is developed to predict the effects of the solar spectrum wavelengths on the PV module performance. The model is characterized and validated under real meteorological conditions where experimental temperature and output power of the PV module measurements are shown to agree with the predicted results. The model is used to validate the concept of active optical filtering. Since this model is wavelength-based, it is used to design an active optical filter for PV applications. Applying this filter to the PV module is expected to increase the output power of the module by filtering the spectrum wavelengths. The active filter performance is optimized, where different cutoff wavelengths are used to maximize the module output power. It is predicted that if the optimized active optical filter is applied to the PV module, the module efficiency is predicted to increase by about 1%. Different technologies are considered for physical implementation of the active optical filter.