This record of study summarizes the work accomplished during the internship at the Energy Systems Laboratory of the Texas Engineering Experiment Station. The internship project was to develop a tool to optimize the building parameters so that the overall building energy cost is minimized. A metaheuristic: genetic algorithm was identified as the solution algorithm and was implemented in the problem under study. Through two case studies, the impacts of the three genetic algorithm parameters, namely population size, crossover and mutation rates, on the algorithm's overall performance are also studied through statistical tests. Through these statistical tests, the optimum combination of above the mentioned parameters is also identified and applied. Finally, a performance analysis based on the case studies show that the tool achieved satisfactory results.

This record of study summarizes the work accomplished during the internship at the Energy Systems Laboratory of the Texas Engineering Experiment Station. The internship project was to develop a tool to optimize the building parameters so that the overall building energy cost is minimized. A metaheuristic: genetic algorithm was identified as the solution algorithm and was implemented in the problem under study. Through two case studies, the impacts of the three genetic algorithm parameters, namely population size, crossover and mutation rates, on the algorithm's overall performance are also studied through statistical tests. Through these statistical tests, the optimum combination of above the mentioned parameters is also identified and applied. Finally, a performance analysis based on the case studies show that the tool achieved satisfactory results.

Renewable marine energy has attracted considerable interest in recent years, especially in the UK due to its excellent location to take advantage of this sustainable energy source. Dierent types of device have been developed over several decades to capture the energy of sea waves but they all need to be able to convert this mechanical energy into electrical energy. The success of wave energy converters (WECs) depends on their eciency, reliability and their ability to react to the variable wave conditions. Although a number of simulation studies have been undertaken, these have used signicantly simplied models and any experimental data is scarce. This work considers a heaving point absorber with a hydraulic power takeo unit. It employs a common hydraulic power take-o design, which uses the heaving motion of the buoy to drive an actuator that behaves like a linear pump. Energy storage is used to provide power smoothing in an attempt to give a constant power output from a hydraulic motor coupled to a generator. Although this design has been presented before, the ineciencies and dynamics of the components have not been investigated in detail. The aim of this work is to create an understanding of the non-linear dynamics of a hydraulic power take-o unit and how these aect the hydrodynamic behaviour of the WEC. A further aim is to predict the eciency of the power take-o unit and determine tuning and control methods which will improve the power generation. In order to do this and test the device in dierent wave conditions, a full hydrodynamic and hydraulic model is developed using the Simulink and SimHydraulics software package. The model is initially tested with regular waves to determine the behaviour of the power take-o unit and a method for adjusting the hydraulic motor displacement depending on the frequency of the incoming wave is investigated. The optimal eective PTO damping to maximise power generation is found to be dependent on the signicant wave frequency and the values of PTO damping are signicantly dierent to previous work using a linear power take-o model which emphasises the importance of including the ineciencies of the hydraulic components. The model is then analysed with irregular waves to predict the behaviour and power levels in realistic wave conditions. Power generation reduces in comparison to regular waves but a similar tuning method to maximise power generation still exists. A hydraulic motor speed control method is shown to increase power generation in irregular waves by maintaining the motor speed within an acceptable working range. Wave data from the Atlantic Ocean is then used to investigate the benets of an adaptive tuning method which uses estimated wave parameters for a number of dierent sea conditions. Results show only minimal gains from using active tuning methods over a passive method. However, results revealed signicant power losses in both calm and rough sea conditions with the PTO most ecient, at approximately 60%, in an average sea power. A scaled experimental power take-o unit is developed to help validate the simulation results. The power take-o unit is tested using a hardware-in-the-loop system in which the hydrodynamic behaviour of the WEC is predicted by a realtime simulation model. The experimental results show good agreement to the simulation with the PTO showing similar characteristics and tuning trends for maximising power generation.

Over the past several decades, memory technologies have exploited continual scaling of CMOS to drastically improve performance and cost. Unfortunately, charge-based memories become unreliable beyond 20 nm feature sizes. A promising alternative is Phase-Change-Memory (PCM) which leverages scalable resistive thermal mechanisms. To realize PCM's potential, a number of challenges, including the limited wear-endurance and costly writes, need to be addressed. This thesis introduces novel methodologies for encoding data on PCM which exploit asymmetries in read/write performance to minimize memory's wear/energy consumption. First, we map the problem to a distance-based graph clustering problem and prove it is NP-hard. Next, we propose two different approaches: an optimal solution based on Integer-Linear-Programming, and an approximately-optimal solution based on Dynamic-Programming. Our methods target both single-level and multi-level cell PCM and provide further optimizations for stochastically-distributed data. We devise a low overhead hardware architecture for the encoder. Evaluations demonstrate significant performance gains of our framework.

In a water distribution system from groundwater supply, the bulk of energy consumption is expended at pump stations. These pumps pressurize the water and transport it from the aquifer to the distribution system and to elevated storage tanks. Each pump in the system has a range of possible operating conditions with varying flow rates, hydraulic head imparted, and hydraulic efficiencies. In this research, the water distribution system of a mid-sized city in a subtropical climate is modeled and optimized in order to minimize the energy usage of its fourteen pumps. A simplified model of the pipes, pumps, and storage tanks is designed using freely-available EPANET hydraulic modeling software. Physical and operational parameters of this model are calibrated against five weeks of observed data using a genetic algorithm to predict storage tank volume given a forecasted system demand. Uncertainty analysis on the calibrated parameters is performed to assess model sensitivity. Finally, the pumping schedule for the system's fourteen pumps is optimized using a genetic algorithm in order to minimize total energy use across a 24-hour period.

The impact of distributed energy resources (DERs), electric vehicles/plug-in hybrid electric vehicles (EVs/PHEVs), and smart appliances on the distribution grid has been expected to be beneficial in terms of environment, economy, and reliability. But, it can be more beneficial by implementing smart controls. In the absence of additional controls, a negative effect was identified regarding the service lifetime of power distribution components. This research presents a new class of a smart house energy management system that can provide management and control of a residential house electric energy without inconvenience to the residents of the house and without overloading the distribution infrastructure. The implementation of these controls requires an infrastructure that continuously monitors the house power system operation, determines the real-time model of the house, computes better operating strategies over a planning period of time, and enables control of house resources. The smart house energy management system provides benefits for the good of utility and customer. In case of variable electricity rates, the management system can reduce the customer’s total energy cost. The benefits can be also extended to provide ancillary services to the utility such as control of peak load and reactive power support– assuming that this is worked out under a certain mutually beneficial arrangement between the utility and customer.

Electric power distribution systems are gradually adopting new advancements in communication, control, measurement, and metering technologies to help realize the evolving concept of Smart Grids. Future distribution systems will facilitate increased and active participation of customers in Demand Side Management activities, with customer load profiles being primarily governed by real-time information such as energy price, emission, and incentive signals from utilities. In such an environment, new mathematical modeling approaches would allow Local Distribution Companies (LDCs) and customers the optimal operation of distribution systems and customer's loads, considering various relevant objectives and constraints. This thesis presents a mathematical model for optimal and real-time operation of distribution systems. Thus, a three-phase Distribution Optimal Power Flow (DOPF) model is proposed, which incorporates comprehensive and realistic models of relevant distribution system components. A novel optimization objective, which minimizes the energy purchased from the external grid while limiting the number of switching operations of control equipment, is considered. A heuristic method is proposed to solve the DOPF model, which is based on a quadratic penalty approach to reduce the computational burden so as to make the solution process suitable for real-time applications. A Genetic Algorithm based solution method is also implemented to compare and benchmark the performance of the proposed heuristic solution method. The results of applying the DOPF model and the solution methods to two distribution systems, i.e., the IEEE 13-node test feeder and a Hydro One distribution feeder, are discussed. The results demonstrate that the proposed three-phase DOPF model and the heuristic solution method may yield some benefits to the LDCs in real-time optimal operation of distribution systems in the context of Smart Grids. This work also presents a mathematical model for optimal and real-time control of customer electricity usage, which can be readily integrated by industrial customers into their Energy Hub Management Systems (EHMSs). An Optimal Industrial Load Management (OILM) model is proposed, which minimizes energy costs and/or demand charges, considering comprehensive models of industrial processes, process interdependencies, storage units, process operating constraints, production requirements, and other relevant constraints. The OILM is integrated with the DOPF model to incorporate operating constraints required by the LDC system operator, thus combining voltage optimization with load control for additional benefits. The OILM model is applied to two industrial customers, i.e., a flour mill and a water pumping facility, and the results demonstrate the benefits to the industrial customers and LDCs that can be obtained by deploying the proposed OILM and three-phase DOPF models in EHMSs, in conjunction with Smart Grid technologies.

In many studies efforts are made to decrease the energy consumption of trains by optimizing their drive style, e.g. accelerate and brake optimally and regenerate electricity when braking. In other studies the goal is to distribute the run time between stations in an optimal way to decrease the energy consumption, given a relatively simple drive style. In this report the goal is to combine these two energy saving methods to obtain as low energy consumption as possible. By coupling one software containing a drive style optimizer with another software which by different optimization methods calculates the optimal run time distribution on a given track this is accomplished. The study also contains a comparison between drive styles, with the goal to find a relatively simple but energy efficient drive style. Finally the dependence between run time distribution and energy consumption is further analysed. The results show that by redistributing the run times the energy consumption can be decreased compared to previously existing time tables. They also show that a relatively simple drive style gives comparable energy consumption compared to the one obtained using a drive style optimizer. Finally the results show that the dependence between run time and energy consumption can be approximated with a simple second order equation.

Wireless sensor networks enable inexpensive distributed monitoring systems that are the backbone of smart cities. In this dissertation, we are interested in wireless sensor networks for traffic monitoring and an emergency flood detection to improve the safety of future cities. To achieve real-time traffic monitoring and emergency flood detection, the system has to be continually operational. Accordingly, an energy source is needed to ensure energy availability at all times. The sun provides for the most inexpensive source of energy, and therefore the energy is provided here by a solar panel working in conjunction with a rechargeable battery. Unlike batteries, solar energy fluctuates spatially and temporally due to the panel orientation, seasonal variation and node location, particularly in cities where buildings cast shadows. Especially, it becomes scarce whenever floods are likely to occur, as the weather tends to be cloudy at such times when the emergency detection system is most needed. These considerations lead to the need for the optimization of the energy of the sensor network, to maximize its sensing performance. In this dissertation, we address the challenges associated with long term outdoor deployments along with providing some solutions to overcome part of these challenges. We then introduce the energy optimization problem, as a distributed greedy approach. Motivated by the flood sensing application, our objective is to maximize the energy margin in the solar powered network at the onset of the high rain event, to maximize the network lifetime. The decentralized scheme will achieve this by optimizing the energy over a time horizon T, taking into account the available and predicted energy over the entire routing path. Having a good energy forecasting scheme can significantly enhance the energy optimization in WSN. Thus, this dissertation proposes a new energy forecasting scheme that is compatible with the platform’s capabilities. This proposed prediction scheme was tested on real data and compared with state-of-theart forecasting schemes on on-node WSN platforms. Finally, to establish the relevance of the aforementioned schemes beyond theoretical formulations and simulations, all proposed protocols and schemes are subjected to hardware implementation.

Indiana University-Purdue University Indianapolis (IUPUI) / Heating, ventilation and air conditioning systems (HVAC) are the single largest consumer of energy in commercial and residential sectors. Minimizing its energy consumption without compromising indoor air quality (IAQ) and thermal comfort would result in environmental and financial benefits. Currently, most buildings still utilize constant air volume (CAV) systems with on/off control to meet the thermal loads. Such systems, without any consideration of occupancy, may ventilate a zone excessively and result in energy waste. Previous studies showed that CO2-based demand-controlled ventilation (DCV) methods are the most widely used strategies to determine the optimal level of supply air volume. However, conventional CO2 mass balanced models do not yield an optimal estimation accuracy. In this study, feed-forward neural network algorithm (FFNN) was proposed to estimate the zone occupancy using CO2 concentrations, observed occupancy data and the zone schedule. The occupancy prediction result was then utilized to optimize supply fan operation of the air handling unit (AHU) associated with the zone. IAQ and thermal comfort standards were also taken into consideration as the active constraints of this optimization. As for the validation, the experiment was carried out in an auditorium located on a university campus. The results revealed that utilizing neural network occupancy estimation model can reduce the daily ventilation energy by 74.2% when compared to the current on/off control.