Master of Science (MS), Ohio University, 2020, Industrial and Systems Engineering (Engineering and Technology)

Process improvements in hospitals usually focus on a single department (eg. emergency department, operating theater, specialty clinic, etc). However, actions taken in one department inevitably affect the performance of other departments. Therefore, higher efficiency improvements can be obtained by considering the patient care process as one synergetic activity involving several departments and various sets of resources. In this research we propose an integrated approach for modeling the patient lifecycle for multiple departments. First we describe a patient flow from his/her entry into the hospital through a progressive care unit until the patient has fully recovered. We use process mapping methods to address value added activities and other necessary activities in the patient lifecycle. Then, a simulation model is developed in Simio using customized objects created in previous works. Those customized objects carry their own logic and behavior. For example, the Bed object includes logic for a patient recovering while using several hospital resources (nurses, therapist) in his/her hospital stay. Those objects were used to build several configurations of an integrated model with multiple departments. Data about patient arrival patterns, their health acuity, and procedure needs were obtained from a real hospital in order to test our approach. The procedures duration data (which were different for different levels of patient acuity and for different surgical and other procedures) were used to obtain service distribution using statistical analysis methods. Modular simulation objects and data distributions from real hospitals allowed us to build an integrated simulation model with several configurations of the process flow. Simulation experiments were performed on these models and performance recorded. The recommendation for implementations in the hospitals is also reported.

Doctor of Philosophy, The Ohio State University, 2015, Computer Science and Engineering

The interactions among fluids and solids create many interesting phenomena that are excessively complex for manual creation in animation. It is popular to model these interactions in physically based simulation but it is challenging especially in real-time applications. Collisions handling is a major bottleneck for solid-solid interaction problems because of high computational cost of neighbor searching in space. Solid-fluid interactions are also difficult to simulate mostly because of the difference in representations of fluids and solids. Typically simulation systems use Eulerian methods for fluids and Lagrangian methods for solids. The most adopted coupling strategy uses solid velocity as boundary condition in fluid solver and integrate fluid pressure along solid boundary to apply force on solid. However, the quality of fluid animation is limited by resolution of Eulerian grid thus it cannot handle interaction with thin features on solids. In this dissertation we focus on specific types of interactions among fluids and solids and develop simulation methods with improved quality and performance toward real-time applications. First we address the problem of cloth, air, and deformable body interactions modeling in a layered structure as commonly seen in real world. We develop an accelerated collision detection method taking advantage of the layer structure to improve efficiency and an accurate anisotropic friction model to achieve fine contact details. The interaction of air and other layers is modeled using a fast air mass field model. Next, we turn to fracture simulation in solid-solid interaction, which is known to be computationally expensive in high resolution. We develop a surface refinement approach that adds fine details to existing low-resolution fracture animation with negligible extra computation cost. Finally, we explore coupling of fluid and solid with thin features. We take a stable and fast approach that couples hybrid Eulerian-Lagrangian fluid and (open full item for complete abstract)

Master of Science in Engineering, Youngstown State University, 2023, Department of Mechanical, Industrial and Manufacturing Engineering

The field of metal additive manufacturing has experienced significant growth in recent years, and Laser Hot Wire Directed Energy Deposition (LHW-DED) has emerged as a popular technology due to its ease of use and ability to produce high-quality metal parts. In this study, we used a nonlinear transient thermo-mechanical coupled finite element model (FEM) in ANSYS APDL to conduct a detailed thermal and structural analysis of the laser hot wire DED metal additive manufacturing process. This analysis aimed to characterize the distortion caused by thermal effects and investigate the transient thermal process. In this study H13 iron chromium alloy material was deposited on an A36 low carbon steel substrate using a bidirectional laser toolpath. To record the temperature profile during printing, we employed a FLIR Infrared (IR) camera, while thermocouples mounted to the base plate measured heat transfer for validation purposes. Post-processing analysis was conducted using the CREAFORM laser 3D scan and Geomagic-X software to measure deformation from the nominal printed geometry. Overall, this study provides a significant contribution to our understanding of laser hot wire DED metal additive manufacturing, which will undoubtedly lead to further advancements in the field. This research has the potential to improve the productivity and quality of the additive manufacture of metals.

Master of Science, The Ohio State University, 2022, Mechanical Engineering

Flexible sheet metal assemblies, such as those seen in automotive bodies, encounter sources of manufacturing variation during their production such as material anisotropy from cold rolling processes, springback behavior following stamping, and distortion after clamping and spot welding due to residual stresses. Designing a product and corresponding manufacturing process which minimizes these variations requires significant iteration of design and analysis, performed by individuals with expert knowledge of complex, multi-stage nonlinear finite element analysis (FEA). The over-arching goal of this research is to investigate the use of machine learning (ML) for predicting manufacturing variations in flexible assemblies, which would significantly reduce these development times. To accomplish this, an automated, multi-stage, explicit FEA workflow for producing a large, balanced, and validated dataset of formed sheet metal components and assemblies has been constructed as a preliminary case study for a larger NSF project. The corresponding dataset has also been used to demonstrate its compatibility for training a machine learning algorithm – namely a fully connected neural network (FCNN) – for use in tooling and process design. To construct the workflow, first a simulation procedure for component forming/springback was set up and verified with the NUMISHEET 1993 U-draw/bending benchmark using implicit FEA. Next, the verified forming/springback simulation was used to investigate whether implicit or explicit FEA would be more appropriate for automated simulations. Following this, methods for transfer of relevant results to succeeding simulation stages were developed. Using these methods, simulations of clamping components and joining with simplified (adiabatic) spot welds were then set up using explicit FEA. Additionally, both variety and balance were introduced to key workflow inputs through parameterization and design of experiments methodologies. The finite element (F (open full item for complete abstract)

Master of Computer Science, Miami University, 2018, Computer Science and Software Engineering

Parallel Discrete Event Simulation (PDES) conducted using emerging shared memory many-core CPUs presents capacity for even greater performance by 1) eliminating the need for message passing and associated serialization/deserialization overheads, and 2) reducing memory requirements by allowing a single copy of an event to be shared between multiple threads. However, the overall performance of a PDES is highly contingent on the speed and capacity of its pending event set data structure. Accordingly, we present a simple, thread-safe priority queue called 3tSkip for managing pending events. Our design takes advantage of contemporary synchronization primitives, including atomics and lock-free data structures to ensure good performance. The priority queue has been incorporated into a redesigned version of a parallel simulator called MUSE, to enable PDES on shared memory platforms. The effectiveness of the proposed solution has been assessed using standard PDES benchmarks. Our analysis identifies many critical design obstacles to multi-threaded design and presents novel solutions to those design obstacles. Our solution achieves significant speedup in high granularity scenarios, when compared to existing MUSE simulator, though more work is required before multithreaded design becomes effective in a broad range of scenarios.

Doctor of Philosophy, The Ohio State University, 2018, Computer Science and Engineering

Subspace Simulation or Model reduction is a technique to simplify simulations of systems modeled by Ordinary Differential Equations. The essential idea is to project the high dimensional sparse equations on to lower dimensional dense equations and then re-projecting the solution back to the original space. Such a method has a much faster computation and lower memory footprint as compared to a full space simulation. Recalculating the subspace basis of a deformable body is a computationally expensive yet mandatory operation. We show that the subspace of a modified body can be efficiently obtained from the subspace of its original version if mesh changes are small. Our basic idea is to approximate the stiffness matrix by its low-frequency component, so we can calculate new linear deformation modes by solving an incremental eigenvalue decomposition problem. We also show a hybridized approach to calculate the modal derivatives and finally we demonstrate that the cubature samples trained for the original mesh can be reused in fast reduced force and stiffness matrix evaluation. Next, we handle the problem of inverse simulation in deformable bodies. We present a novel system for interactive elastic shape design in both forward and inverse fashions. Using this system, the user can choose to edit the rest shape or the quasistatic shape of an elastic solid and obtain the other shape that matches under the quasistatic equilibrium condition at the same time. The development of this system is based on the discovery that inverse quasistatic simulation can be immediately solved by Newton's method with a direct solver. To implement our simulator, we propose a Jacobian matrix evaluation scheme for the inverse elastic problem and we present step length and matrix evaluation techniques that improve the simulation performance. While our simulator is efficient, it is still not fast enough for the system to generate the result in real time. Our solution is a shape initialization met (open full item for complete abstract)

Doctor of Philosophy (PhD), Ohio University, 2016, Instructional Technology (Education)

Nursing programs across the country are facing the challenge of providing students with optimal clinical skill learning experiences. Faculty members are rising to this challenge in a variety of ways, and current research in nursing education advocates the addition of interprofessional education to the nursing curriculum. History and governing bodies explain how nursing education can achieve this goal but faculty have to implement required and suggested changes. Further research is needed to gain insight into perceptions of the baccalaureate nursing faculty of interprofessional education and how the simulation in interprofessional experience contributes to overall patient safety. Patient safety is a priority in nursing and by using simulation as a safe place for a nursing student to learn, and then patient safety is enhanced. A qualitative case study was used to explore how the use of interprofessional education by baccalaureate nursing faculty can prepare nursing students to provide quality care to real patients. Baccalaureate nursing faculty from a large midwestern university in the United States of America participated in a case study that involved an interprofessional education simulation experience. Faculty were asked to participate in a set of interviews both before and after the completion of a simulated interprofessional education experience. Additionally, the case study participants were asked to provide narrative reflective journal writings of their experiences with any simulation interprofessional education experience or any other interprofessional experience. Additional methods of observation, field notes, and focus group, were used by the researcher to gain an understanding of the faculty lived experiences. Findings from this study will aid in curricular changes within healthcare education regarding simulation in interprofessional education.

Doctor of Philosophy, Case Western Reserve University, 2011, Macromolecular Science and Engineering

We studied the self-assembly dynamics of two polymeric systems, block copolymer micelles and supramolecular polymer solutions using computer simulation. Dissipative Particle Dynamics simulations were applied to study the equilibrium properties, kinetics of micellization and equilibrium chain-exchange in A2B3 and A4Bx(x=4,6,8) diblock copolymer micelle solutions. The critical micelle concentration, micelle aggregation number distribution and micelle structure were found to agree well with previous experimental and theoretical studies. The time-evolution of micelles from unimers is found to follow three stages: unimer consumption, equilibration of the number of micelles progressing mainly by the fusion/fission mechanism and slow adjustment of the weight-average aggregation number by micelle fusion, unimer and small aggregate exchange. The effect of polymer concentration, hydrophobic interaction energy and block length on the kinetics of micellization were also considered. By performing micelle hybridization simulations, we found the equilibrium chain exchange follows a first-order kinetic process and the characteristic time, mainly determined by chain expulsion and does not depend on polymer concentration. The chain exchange characteristic time, τ, increases exponentially with core block length, NA and interaction parameter between blocks, χAB as τ ~ exp(0.67χABNA). We also found that in contrast to theoretical predictions, chain exchange between micelles occurs more rapidly for micelles with a longer corona-block length due to a higher compatibility of diblock copolymers and therefore a lower potential barrier for chain expulsion. Using coarse-grained molecular dynamics simulations we studied the equilibrium and rheological properties of dilute and semi-dilute solutions of head-to-tail associating supramolecular polymers with our newly-developed model for spontaneous reversible association. We found that for a given spacer length all shear-rate-dependent reduced visc (open full item for complete abstract)

Doctor of Philosophy (PhD), Ohio University, 2024, Civil Engineering (Engineering and Technology)

This research project is intended to prepare an integrated model that can be used as a toolkit for Kabul City in Afghanistan and later for any geographical area to help mitigate the damage due to flooding during the peak-flow period of surface water flow, and also contribute to groundwater recharge as a tool to assist during the drought eras that the city faces. In this project, the surface flow of water over a course of time is assessed, with regards to the volume of flow and the capacity of the water bodies holding it such as river, canals, waterways etc. After the assessment of the volume of flow, this study project proposes areas for water retention during the peak flow period that can provide a sustainable water storage solution, consequently that can contribute to the recharge of the groundwater table for that particular geographical zone as well. The selection of water retention zone is conducted based on the topographical assessment, availability of land, possibility of groundwork and construction and other related factors. In this project, the surface flow of water inside and around Kabul City was analyzed, and accordingly the peak flow periods during the last few decades were examined, which had caused floods and lots of damages inside the city as well. Based on the topographical assessment, land use plan and peak flow simulations, two major sites were selected as water retention zones. These zones were intended to reduce the intensity of water during the peak flow period, retain the water, help mitigate the damage due to the floodings and also contribute in the recharge of groundwater. The wells were established in the selected sites to monitor the groundwater table. The surface flow for the city was simulated to understand its impact on the two selected sites. Based on various mathematical equations, the groundwater recharge was estimated and compared with the actual groundwater table over the time when the wells were monitored. Thi (open full item for complete abstract)

Ph.D., Antioch University, 2024, Antioch Seattle: Counselor Education & Supervision

This multilevel mixed methods investigation examines the experiences of developing self-evaluation skills for simulation fellows in an interprofessional simulation fellowship program. Interprofessional fellows (N = 12) and faculty (N = 4) engaged in a three-phase study using video-assisted learning tools to explore the differences in self-evaluation (perceived performance) and faculty evaluation (actual performance) in developing debriefing skills. For the quantitative component, fellows and faculty completed the DASH© tool to evaluate the quality of debriefing to help close the gaps between fellow self-evaluation and faculty evaluation. For the qualitative component, video-stimulated think-aloud and video-assisted debriefing the debriefer were utilized to understand how video-assisted learning tools contribute to self-evaluation skills from each level separately. A combined focus group and thematic analysis were utilized to identify facilitators and barriers to self-evaluation. Integrative case examples of developing self-evaluation skills are restoried for mixed analysis and data integration. Through which, meta-inferences are drawn out to understand the experiences, interactions, and mechanisms of the multilevel phenomenon. The results indicated that fellows generally overestimate their skills performance, and video-assisted learning tools support in the development of more realistic self-perception eliminating underestimation and closing the gap between perceived and actual performance. A model for interprofessional collaboration is proposed for scaffolded feedback practices to promote self-evaluation of skills and performance. Implications for counselor education, healthcare simulation, and health professions education are presented.

Master of Science (MS), Ohio University, 2023, Industrial and Systems Engineering (Engineering and Technology)

Using computer simulation in healthcare is a longstanding endeavor, where alternative scenarios and system configurations can be tested before they are implemented in the real world. One facet of healthcare that has an apparent lack of simulation work is that of emergency transportation, especially that which has a goal to reduce the amount of time a crew works over their designated shift. In collaboration with a Midwest USA Hospital, the researcher sought to model the current system, then experiment by altering crew schedule start times, the number of crews, and policy to reduce the frequency and duration a crew works past their scheduled end time. Therefrom, a framework was developed to help other institutions with similar aims. After the simulation model was verified and validated, experiments varying the start times of the crews, scheduling the crew constituent resources separately, and the number of crews were investigated. With the goal to reduce the time a crew spends working past their scheduled shift without negatively affecting productivity, scenarios that performed similarly in the number of transport requests serviced where there was a decrease in over-shift metrics were identified. The results of the simulation experimentation can be summarized this way: reducing over-shift comes at a cost, likely, either a reduction in productivity or an additional crew. A seven-themed framework for future studies was derived: a sound mission, system understanding, data availability/understanding, respect for process, simulation experience, results analysis and recommendations, and empathy.

Doctor of Philosophy (PhD), Ohio University, 2022, Physics and Astronomy (Arts and Sciences)

Despite the long standing history of the research, production, and application of amorphous and glassy materials, generating good quality models still remains a challenge. The challenge arises from the inherent lack of the long range order, characteristic of crystals, in amorphous materials. Researchers have developed various techniques to create models of amorphous materials ranging from random Monte Carlo to classical molecular dynamics and from ab initio to the most recent machine-learned methods. In this dissertation, we apply force enhanced atomic refinement (FEAR) whereby experimental information from diffraction measurements are used jointly with ab initio density functional theory (DFT) based energy minimization to produce models of various amorphous materials that agree with diffraction data and are a suitable energy minimum of the chosen interatomic potential functions. By generating models of metal oxides and chalcogenides, we show that this method is broadly applicable to amorphous material if the experimental diffraction data is available. We used this to study the annealing induced changes in the structure of ZrO2-Ta2O5, a potential candidate for mirror coatings for the Laser Interferometer Gravitational-wave Observatory (LIGO) interferometer mirrors. We find that annealing increases the fraction of corner-shared metal-oxygen polyhedra in this material. Motivated by interest in carbonaceous materials, we studied the graphitization of carbon at temperatures near 3000 K. For the first time, we accurately simulate the process of graphitization and the mechanisms of layering. We have seen that individual layers of amorphous graphite are topologically disordered with some pentagon and heptagon carbon rings and have studied the effects of this disorder on charge density distribution, electronic density of states, and electronic conduction. The study of carbonaceous materials was extended to study the reactivity of carbon surfaces to diff (open full item for complete abstract)

Master of Science (MS), Wright State University, 2022, Computer Science

Healthcare organizations attract a diversity of caregivers and patients by providing essential care. While interacting with people of various races, ethnicity, and economical background, caregivers need to be empathetic and compassionate. Proper training and exposure are needed to understand the patient's background and handle different situations and provide the best care for the patient. With social determinants of health (SDOH) as the basis, the thesis focuses on providing exposure through “Wright LIFE (Lifelike Immersion for Equity) - A simulation-based training tool” to two such scenarios covering patients from the LGBTQIA+ community & autism spectrum disorder (ASD). This interactive tool helps to create mindfulness about the social and economic disparities faced by the patients through realistic and captivating gameplay. Though the primary focus of the “Wright LIFE” application is “Digital Learning”, it would help to understand how effective the application is in terms of improving the provider's abilities. Through statistical evidence, the tool can be improved, which in turn will improve the user experience. For this analysis, during the simulation, we also focus on collecting the data gathered from the participants through surveys. The simulation includes different questionnaires where participants can provide feedback at various stages within the simulation. This then allows for a comparison between the participants' responses to see the rate of improvement as a result of the simulation. To analyze the data from the participant's responses, data analysis, and visualization tools help to represent the data using charts, infographics, animations, and many more to assist this in this analytic process. The analysis of the data can help to understand the trend of the participants' responses to the questionnaire. The goal of the questionnaire is to collect participants' responses to assess anxiety, frustration, and compassion levels pre- and post-simulation. A (open full item for complete abstract)

PHD, Kent State University, 2022, College of Education, Health and Human Services / School of Teaching, Learning and Curriculum Studies

Research literature provides evidence that new graduate nurses are often deficient in clinical judgment (CJ). One way to increase CJ is by using simulations. However, the literature is replete with descriptions of the high anxiety that simulation triggers. It is not currently known how anxiety in simulation affects clinical judgment for undergraduate nursing students. Therefore, the purpose of this study was to explore the effect of different types of anxiety on the clinical judgment of undergraduate nursing students in simulation. This research project used a one-group repeated measures quantitative design to answer the research questions using the conceptual framework of Tanner's (2006) model of clinical judgment. A convenience sample of 45 sophomore-level undergraduate nursing students participated in a study to explore how state and trait anxiety impacted their clinical judgment within an introductory simulation. The results indicated that anxiety did not have a significant impact on clinical judgment. When controlling for baseline state and trait anxiety, pre-simulation anxiety level did not significantly predict scores on the Lasater Clinical Judgment Rubric (LCJR) within the simulation. State anxiety did change significantly between the three time measurements, going up to significantly high levels at pre-simulation. These anxiety levels remained high at post-simulation. The findings imply a changed focus to reframe how anxiety is thought about and its effects. Some anxiety is good and facilitative, and therefore, faculty should not be so worried about reducing anxiety for all students. Rather, nursing educators should help students function despite anxiety, in order to prepare them for real world nursing practice.

Master of Science (MS), Wright State University, 2020, Physics

A one-dimensional kinetic particle-in-cell (PIC) MATLAB simulation was created to demonstrate the time-evolution of various plasma distributions. Building on previous plasma PIC programs written in FORTRAN and Python, this work recreates the computational and diagnostic tools of these packages in a more user- and educational-friendly development environment. Plasma quantities such as plasma frequency and species charge-mass ratios are arbitrarily defined. A one-dimensional spatial environment is defined by total length and number and size of spatial grid points. In the first time-step, charged particles are given initial positions and velocities on a spatial grid. After initialization, the program solves for the electrostatic Poisson equation at each time step to compute the force acting on each particle. Using the calculated force on each particle and the “leap-frog” method, the particle positions and velocities are updated and the motion is tracked in phase-space. Modifying parameters such as spatial perturbation, number of particles, and charge-mass ratio of each species, the time-evolution for various distributions are examined. The simulated distributions examined are categorized as the following: Cold Electron Stream, Electron Plasma Waves, Two-Stream Electron Instability, Landau Damping, and Beam-Plasma. The time evolution of the plasma distributions was studied by several methods. Tracking the electric field, charge density and particle velocities through each time step yields insight into the oscillations and wave propagation associated with each distribution. One key diagnostic missing from the original FORTRAN code was the electric field dispersion relation. The numerical dispersion relation allows for further insight into modelling plasma oscillations/waves in addition to the kinetic/field energies and electric field tracking present in the original code. Simulated results show agreement with other kinetic simulations as well as plasma theory.

Doctor of Philosophy, University of Akron, 2020, Polymer Engineering

The nature of glassy aging has been a topic of study for over half a century, and yet a number of open questions remain in the understanding of the glassy state. Since a polymer's physical and mechanical properties are directly dependent on its molecular structure and changes in that structure alter the physical properties of the glass, considerable economic impact can result from aging-related physical changes. Characterization of aging dynamics in under-dense and over-dense glasses and a comparison of the aging response in solvent-processed vs thermally-quenched glasses are two important questions that are addressed here. This work reports on the development of a protocol for studying physical aging via molecular dynamics simulation after a near-instantaneous temperature quench. The resulting data display characteristic experimental signatures of glassy aging in both a pure polymer and a polymer-plasticizer system, indicating that this protocol can potentially be used to study aging in a variety of systems. Results indicate that aging dynamics in under-dense and over-dense glasses are fundamentally different in character. Unlike in under-dense glasses, translational dynamics in over-dense glasses are mechanistically different than relaxation in equilibrium glass-forming liquids, which is supported by the finding that relaxation in over-dense glasses occurs through an explosive burst of superdiffusive motion. Addition of a plasticizer appears to moderate this response compared to that of the pure polymer system, which can be attributed to a decrease in system fragility in the plasticized system. Higher additive loadings may have an even greater effect and further research would be beneficial in clarifying this. Aging relaxation time in over-dense glasses obeys a zero parameter dependence on purely equilibrium properties. This finding enables prediction of non-equilibrium relaxation time given knowledge only of the starting temperature and the in-equilibrium (open full item for complete abstract)

Master of Science, The Ohio State University, 2019, Dentistry

Objective: In the event of a medical emergency in the dental office, the supervising dentist must be able to effectively lead the office team in a concerted effort to stabilize the patient and transfer them to higher-level medical care. This study investigates the impact of a simulation-based medical emergency training curriculum on the ability of general practice residents to effectively manage medical emergencies in a dental environment. Methods: An interventional and pre-post educational trial of 16 general practice residency participants was carried out at The Ohio State University College of Dentistry. Eight participants completed a standard training curriculum as the control group and eight completed a modified training curriculum as the intervention group. The intervention consisted of a simulation-based education curriculum designed for dentists called Medical Emergency Management in the Dental Office (MEMDO). Near the completion of residency, each participant experienced a summative performance-based assessment using an Objective Structured Clinical Examination (OSCE), which was later reviewed and scored by a customized 128-point scoring grid. Additionally, the intervention group completed a baseline performance assessment at the beginning of their residency. Four calibrated faculty reviewers scored each OSCE independently. These data were subsequently analyzed using nonparametric statistical tests with alpha set to 0.05. Reviewer consistency was assessed by calculating an intraclasscorrelation coefficient. All participants completed a survey of demographic information and 11 Likert-type questions. Results: The intervention group performed significantly better than the control group (p=0.0009). This group improved their post-intervention score by an average of 36.9 points out of 128. The intraclass correlation coefficient was found to be 0.9795. Surveys found all participants in agreement with the importance of medical emergency preparedness of all de (open full item for complete abstract)

Master of Sciences, Case Western Reserve University, 2019, EECS - Computer and Information Sciences

Computer vision tasks require collecting large volumes of data, which can be a time consuming effort. Automating the collection process with simulations speeds up the process, at the cost of the virtual data not closely matching the physical data. Building upon a previous attempt to bridge this gap, this thesis proposes three nuances to improve the correspondence between simulated and physical 3-D point clouds and depth images. First, the same CAD files used for simulated data acquisition are also used to 3-D print physical models used for physical data acquisition. Second, a new projection method is developed to make better use of all information provided by the depth camera. Finally, all projection parameters are unified to prevent the deep learning model from developing a dependence on intensity scaling. A convolutional neural network is trained on the simulated data and evaluated on the physical data to determine the model's generalization ability.

Doctor of Philosophy, The Ohio State University, 2018, Mechanical Engineering

Hybrid electric vehicles are a result of a global push towards cleaner and fuel-efficient vehicles. They use both electrical and traditional fossil-fuel based energy sources, which makes them ideal for the transition towards much cleaner electric vehicles. A key part of the hybridization effort is designing effective energy management algorithms because they are crucial in reducing fuel consumption and emission of the hybrid vehicle. In the automotive industry, energy management systems are designed, prototyped, and validated in a software simulation environment before implementation on the hybrid vehicle. The software simulation uses model-based design techniques which reduce development time and cost. Traditionally, the design of energy management systems is based on statutory drive-cycles. Drive-cycle based solutions to energy management systems improve fuel economy of the vehicle and are well suited for statutory certification of fuel economy and emissions. In recent times however, the fuel economy and emissions over real-world driving is being considered increasingly for statutory certification. In light of these developments, methodologies to simulate and design new energy management strategies for real-world driving are needed. The work presented in this dissertation systematically addresses the challenges faced in the development of such a methodology. This work identifies and solves three sub-problems which together form the methodology for model-based real-world look-ahead energy management system development. First, a simulation framework to simulate real-world driving and look-ahead sensor emulation is developed. The simulation framework includes traffic simulation and powertrain simulation capabilities. It is termed traffic integrated powertrain co-simulation. Second, a comprehensive algorithm is developed to utilize look-ahead sensor data to accurately predict the vehicle's future velocity trajectories. Finally, through the use of optimal c (open full item for complete abstract)

PhD, University of Cincinnati, 2018, Engineering and Applied Science: Computer Science and Engineering

Forward-time population genetic simulators are computational tools used in the study of population genetics. Simulations aim to evolve the genetic state of a population relative to a set of genetic models that reflect the processes that occur in nature under various configurations. Often, these simulations are limited to evolutionary scales that can be represented within the memory space and feasibly computed using a standard workstation computer. This presents a general challenge of how to represent the genetics of a population to enable evolutionary scenarios of sufficient scale to be performed on a memory constrained system. In addition, as the evolutionary scales increase so too does the computational time necessary to complete the simulation. This work considers the general problems of scale and performance as they relate to forward-time population genetic simulation. It explores the representation of a population from the perspective of a graph. Improved memory utilization and computational performance are achieved through the use of a binary adjacency matrix representation of the graph. This use of this representation is generally uncommon in forward-time population genetic simulation. Further improvements are made to the performance of the simulator through the use of parallel computation. This work considers a forward-time population genetic simulation from both a taskand a data- parallel perspective. Each of these perspectives present certain challenges and offer different levels of performance gains. The utilization of the binary adjacency matrix representation enables each of these parallel approaches to be achieved. Finally, although scale and performance improvements are enabled through the use of a binary adjacency matrix representation of the graph, it does have limits in forward-time population genetic simulation. These limits are related to the density of the graph. This work offers a situation where this representation w (open full item for complete abstract)