41530 research outputs found
Sort by
Design and Implementation of a Prototype Pendulum Platform for Aerospace Controls Education
Get PDFThis thesis presents a modular, low-cost, prototype inverted pendulum platform for undergraduate Aerospace control theory research and education. The platform\u27s default configuration integrates a propeller and reaction wheel as its actuators, with multiple sensors available to measure the pendulum\u27s states. The platform utilizes open-source software and hardware to power the actuators, and uses MATLAB/Simulink to streamline data collection. System identification techniques are applied to the platform, and the mathematical model of the pendulum is verified. Momentum exchange between the reaction wheel and propeller during closed-loop stabilization of the pendulum is demonstrated
Neural Compression Inference Accelerator: a Cost and Energy-Effective Alternative to Conventional Machine Learning Inference Methods
Get PDFRecent developments in machine learning and artificial intelligence have sparked an influx of workloads that require specialized computer hardware for cloud services. The hardware running machine learning models predominantly consists of graphics processing units (GPUs) and tensor processing units (TPUs). However, these com- ponents are expensive for cloud services to purchase, costly for customers to rent, prone to price spikes, and energy-intensive. In this research we show that both cloud services and customers would benefit from utilizing field programmable gate arrays (FPGAs) to alleviate the aforementioned challenges. An FPGA can be configured as a machine learning accelerator, operating similarly to a GPU or TPU. We propose an FPGA-based architecture that utilizes compressed neural networks. This provides an alternative hardware option for those running machine learning inference, resulting in savings in server construction costs, power consumption, workload distribution, and providing a more affordable option for customers. We evaluate the system using architectural simulations and cloud-based deployment and demonstrate 76.2% cost and 26.5% energy savings on average from using the Neural Compression Inference Accelerator (NCIA)
Composite Solid Rocket Motor Case Optimization With Matlab Genetic Algorithm
No full textThis study utilizes MATLAB genetic algorithm (GA) to perform optimization on a solid rocket motor (SRM) chamber structure. The (SRM) structure is modelled as a cylindrical composite laminate where unidirectional fiber orientations are the primary design variable in the optimization problem formulation. Additional design variables such as chamber length, diameter, and internal pressure are altered along with the ply angles to minimize the safety margin towards a near-zero value and encourage efficient layup design. Classical lamination theory with Tsai-Wu failure criterion are used to calculate ply-level safety margins. A thickness minimization scheme is employed to reduce the ply count from a conservative predetermined value to assist in limiting the structural mass of the system. The optimized layup configurations produced by the MATLAB GA implementation are compared to a previous optimization study from literature for mass, dimensions, and propulsion capabilities. Finite element analysis is used as a validation tool to confirm the safety of the proposed design. Finally, the capabilities of the GA are evaluated against a commercially available structural optimization software package on the same SRM chamber model
Compact Flexible Heart Rate Monitor Simulations
No full textCardiovascular disease is a global health concern; heart rate monitors can help assess cardiac health. Sensors must be effective and non-invasive. This project develops and characterizes flexible substrate antennas for heart beat detection through reflection coefficient measurements. Antenna substrate printing and sensor applications are through external collaborations. The project includes curved ground plane and substrate HFSS models to determine reflection coefficient responses to chest expansions
Filters for Fiji: A Modified Cracker Box Biosand Filter for Rural Fiji
Get PDFWater quality remains a major issue throughout rural Fiji, as current solutions rely on outside organizations providing money or technology. The goal of Filters for Fiji’s research was to develop a sustainable water filtration method for the people of rural Fiji, through ethnographic and biological research. In order to be sustainable, the filter had to be inexpensive (less than USD$15), simply built and maintained, and suitable for the end-user. The driving question of the research was: can a plastic, modified biosand filter be constructed from materials that could be sourced locally in rural Fijian communities? To answer this question, feedback from focus groups conducted with rural Fijians, the potential end-users, was used to re-design an existing biosand filter prototype. The previous prototype was in development by an Engineers without Borders (EWB) project focused on water quality and quantity in rural Fiji. Mechanical testing was conducted to allow for iteration that reduced leakage and improved flow rate of the new design. Then, two filter prototypes were tested over 59 days for biological removal efficiency. Drumm Reservoir water, inoculated with K-12 E. coli, was filtered by the prototypes, while the turbidity, total coliform, and E. coli levels were recorded for the filters’ influent and effluent. Analysis of the testing results found the design to remove approximately 60% of E. coli and turbidity on average, which yields significant benefits to the thousands of Fijians who currently lack access to potable water but does not meet CAWST (Center for Affordable Water and Sanitation) biosand filter expectations. Since the prototypes do not meet CAWST expectations the design cannot be recommended for implementation currently. Although the biological testing did not yield perfect results, the ethnographic and lab research completed for this project led to the finding that: “Community centered design is essential for engineering sustainable water filtration solutions in majority world countries”. With the data collected, Filters for Fiji intends to publish this research as a free technology to allow for continued iteration to produce an even more effective human-centered water filtration system for the people of rural Fiji and encourage community centered design in engineered filtration systems
Band Motor Crankshaft for Hydrogen Combustion Engine
Get PDFThis report outlines the steps in which a group of four mechanical engineers at California Polytechnic State University California, San Luis Obispo, addressed an issue brought up by Professor Roger Benham from the Materials Department. The task included developing an automated and modular testing platform for a revolutionary concept crankshaft prototype – meant for a hydrogen combustion engine - developed and produced by the sponsor: Roger Benham. A testing platform is necessary to measure the performance of the shaft design created to eventually implement this subsystem into a complete hydrogen engine design. Values being tested include geometric validation, maximum static loading, maximum dynamic loading, and fatigue testing – these values are calculated via a physical strain gauge on the shaft and a separate Microsoft Excel calculator for validation. Throughout this process, we have implemented the standard engineering design process along with its corresponding reports alongside other specific techniques such as finite element analysis and material design analysis. Overall, the team was able to successfully build an automated test platform that loaded the crankshaft but fell short on the load requirements. Although desired motion was achieved, our system fell short of all load requirements. The system did not put enough torque onto the shaft for our strain gauges to pick up any deflection or torsion. Consequently, no stress values were able to be read. In summary, our team was able to successfully implement an automated and modular design that adequately simulates the desired motion of the crankshaft but fails to output stress values due to lack of torque
Feminist Pedagogy in the STEM Research Laboratory: an Intersectional Approach
Get PDFThe research laboratory is a crucial and indispensable classroom for STEM education. It is where we practice science as a craft and test the ideas that awaken our curiosity, allowing us to create knowledge. It is also a space where challenges await and struggles are imminent. Thus, supporting mentees through their traineeship in a research lab requires an intersectional approach and lens to provide equitable mentorship and guidance. The concept of intersectionality, initially devised by Black feminist professor Kimberlé W. Crenshaw, can be employed to generate practices and frameworks that democratize laboratory culture and provide trainees with a space in which they shape the laboratory culture while helping them recognize their positionality. This critical commentary provides insights and experiences when incorporating feminist frameworks to sustain equitable working environments in the research lab setting, specifically in leading a research group composed predominantly of Puerto Rican women
Ungrading: Reflections Through a Feminist Pedagogical Lens
Get PDFUngrading is a pedagogical approach in which no grades are given on any assignments. Instead, students are provided with many opportunities to submit work and gain feedback. The goal is to shift student focus from achieving a grade to growth as a learner and a person. As instructors, our ungrading approach utilized personalized learning plans, checkpoint reflections, and student-professor learning conferences to put agency in the hands of our students. We employed this method in upper-level biology and computer science courses and provide critical reflections here regarding our experiences and the connections between this approach and feminist STEM pedagogy tenets. We were impressed with students’ self-reported feelings of engagement with the material, enthusiasm for risk-taking, and reduced stress. Feminist themes of reducing the classroom power gap, active student participation, and addressing and disrupting systems of oppression are met through student ownership of their learning and challenging of traditional power dynamics between the student and professor
Exploring representation in Microbiology introductory courses can encourage a more inclusive and inspiring environment for students and instructors
Get PDFMicrobiology has a relatively brief history where significant discoveries are often linked with major events in human history - from disease outbreak to industrialization to climate change. The founders of key microbiological principles span across continents, genders, ethnicities, and socioeconomic status. However, the portrait described in many introductory textbooks center around a lone, typically white male scientist. Such narratives not only are misleading regarding the development of key principles in microbiology but can also reinforce inappropriate stereotypes as to whom belongs in microbiology. In our introductory microbiology course, we designed group work for Zoom break-out rooms to help engage students in the online environment. Here we reflect upon the benefits of this activity in enhancing not only the experience for our students but also that of the instructors