Computer-Aided Engineering and Design Commons^™

Addressing Urban Traffic Congestion: A Deep Reinforcement Learning-Based Approach, Tairan Liu

Mineta Transportation Institute

In an innovative venture, the research team embarked on a mission to redefine urban traffic flow by introducing an automated way to manage traffic light timings. This project integrates two critical technologies, Deep Q-Networks (DQN) and Auto-encoders, into reinforcement learning, with the goal of making traffic smoother and reducing the all-too-common road congestion in simulated city environments. Deep Q-Networks (DQN) are a form of reinforcement learning algorithms that learns the best actions to take in various situations through trial and error. Auto-encoders, on the other hand, are tools that help simplify complex data, making it easier for the DQN to …

Multi-Objective Design Optimization Of Hypoid Geared Rotor Systems, Xinqi Wei

Mechanical and Aerospace Engineering Dissertations - Archive

Hypoid gears represent one of the most generalized and complex forms of gearing, widely used for power transmission of skew shafts in vehicles, aviation, and marine transmission applications. Optimizing their performance remains challenging due to the complex tooth surface and contact behavior. Specifically, the design parameters of the tooth surface are multi-scale, interdependent, and subject to strong constraints, leading to strong nonlinearity and an ill-conditioned Jacobian matrix in the parameter identification model. Moreover, feasible and insensitive contact conditions are difficult to constrain due to the inherent complexity of local conjugate contact between the meshing surfaces. These challenges significantly increase optimization …

Deep Reinforcement Learning For The Design Of Structural Topologies, Nathan Brown

All Dissertations

Advances in machine learning algorithms and increased computational efficiencies have given engineers new capabilities and tools for engineering design. The presented work investigates using deep reinforcement learning (DRL), a subset of deep machine learning that teaches an agent to complete a task through accumulating experiences in an interactive environment, to design 2D structural topologies. Three unique structural topology design problems are investigated to validate DRL as a practical design automation tool to produce high-performing designs in structural topology domains.

The first design problem attempts to find a gradient-free alternative to solving the compliance minimization topology optimization problem. In the proposed …

Development Of A Reverse Engineered, Parameterized, And Structurally Validated Computational Model To Identify Design Parameters That Influence American Football Faceguard Performance, William Ferriell

All Dissertations

Traumatic brain injury (TBI) continues to have the greatest incidence among athletes participating in American football. The headgear design research community has focused on developing accurate computational and experimental analysis techniques to better assess the ability of headgear technology to attenuate impacts and protect athletes from TBI. Despite efforts to innovate the headgear system, minimal progress has been made to innovate the faceguard. Although the faceguard is not the primary component of the headgear system that contributes to impact attenuation, faceguard performance metrics, such as weight, structural stiffness, and visual field occlusions, have been linked to athlete safety. To improve …

Complete Neighbourhood Search Heuristic Algorithm For Portfolio Optimization, Collether John

Tanzania Journal of Engineering and Technology (TJET)

In portfolio optimization, the fundamental goal of an investor is to optimally allocate investments between different assets. Mean-variance optimization methods make unrealistic assumptions to solve the problem of optimal allocation. On the other hand, when realistic constraints like holding size and cardinality are introduced it leads to optimal asset allocation which differ from the mean variance optimization. The resulting optimization problem become quite complex as it exhibits multiple local extrema and discontinuities. Heuristic algorithms work well for the complex problem. Therefore, a heuristic algorithm is developed which is based on hill climbing complete (HC-C). It is utilized to solve the …

Design And Optimization Of Dynamic System For A One-Kw Free Piston Linear Engine Alternator-Gensets Program, Nima Zamani Meymian

Graduate Theses, Dissertations, and Problem Reports (ETD)

In power/energy systems, free-piston linear machines are referred to as a mechanism where the constrained crank motion is eliminated and replaced with free reciprocating piston motion. Depending on the application, the piston motion can be converted into other types of energy and includes compressed air/fluid, electricity, and high temperature/pressure gas. A research group at West Virginia University developed a free-piston linear engine alternator (LEA) in 1998 and have achieved significant accomplishment in the performance enhancement of the LEAs to date. The present LEA design incorporates flexure springs as energy restoration components and as bearing supports. The advantages of using flexure …

Design And Process Of 3d-Printed Parts Using Composite Theory, Jordan Garcia

Theses and Dissertations--Mechanical and Aerospace Engineering

3D printing is a revolutionary manufacturing method that allows the productions of engineering parts almost directly from modeling software on a computer. With 3D printing technology, future manufacturing could become vastly efficient. However, it is observed that the procedures used in 3D printing differ substantially among the printers and from those used in conventional manufacturing. In this thesis, the mechanical properties of engineering products fabricated by 3D printing were comprehensively evaluated and then compared with those made by conventional manufacturing. Three open-source 3D printers, i.e., the Flash Forge Dreamer, the Tevo Tornado, and the Prusa, were used to fabricate the …

A Parametric Investigation And Optimization Of A Cylindrical Explosive Charge, Logan Ellsworth Beaver

Master's Theses (2009 -)

Explosive device design has a wide impact in the space, manufacturing, military, and mining industries. As a step toward computer assisted design of explosives, an optimization framework was developed using the Design Analysis Kit for Optimization and Terrascale Applications (Dakota). This software was coupled with the hydrocode CTH. This framework was applied to three exploding cylinder models, two in 1D and one in 2D. Gradient descent, dividing rectangles, and a genetic algorithm were each applied to the one-dimensional models. Parametric studies were performed as a basis for comparison with the optimization algorithms, as well as qualifying the 1D model's accuracy. …

Corgi: Compute Oriented Recumbent Generation Infrastructure, Christopher Allen Hunt

Master's Theses

Creating a bicycle with a rideable geometry is more complicated than it may appear, with today’s mainstay designs having evolved through years of iteration. This slow evolution coupled with the bicycle’s intricate mechanical system has lead most builders to base their new geometries off of previous work rather than expand into new design spaces. This crutch can lead to slow bicycle iteration rates, often causing bicycles to all look about the same. To combat this, several bicycle design models have been created over the years, with each attempting to define a bicycle’s handling characteristics given its physical geometry. However, these …

Computer-Aided Design Of Algorithms Of Pulsed Control Of Arc Welding Process Based On Numerical Simulation, Oksana I. Shpigunova, Anatoliy A. Glazunov

The 8th International Conference on Physical and Numerical Simulation of Materials Processing

No abstract provided.

Optimization Of Formula Sae Electric Vehicle Frame With Finite Element Analysis, Alexander Prorok

Williams Honors College, Honors Research Projects

Optimization of Formula SAE Electric Vehicle Frame with Finite Element Analysis

Finite Element Analysis And Topology Optimization Of Differential Case And Control Arm For Static And Fatigue Loading, Pankaj N. Kalan

Dissertations, Master's Theses and Master's Reports

Optimization of the automobile components can result in a significant decrease in vehicle weight, increase in fuel efficiency and reduction in environmental damage. For example, the lightweight vehicle production will save over 4.4×109 GJ of energy and 4.1×108 tons of CO2 over a 10 year period, compared to the current non-light weight vehicles. This equates to 890 billion miles or 42 billion gallons of gas savings (Wu 2016). This study evaluates the weight reduction possible by using manual and automatic (using Optistruct software) topology optimization processes. Optimization of a vehicle differential case done in this study shows that a weight …

Reliability-Based Design Optimization Of Concrete Flexural Members Reinforced With Ductile Frp Bars, Bashar Behnam, Christopher D. Eamon

Civil and Environmental Engineering Faculty Research Publications

In recent years, ductile hybrid FRP (DHFRP) bars have been developed for use as tensile reinforcement. However, initial material costs regain high, and it is difficult to simultaneously meet strength, stiffness, ductility, and reliability demands. In this study, a reliability-based design optimization (RBDO) is conducted to determine minimum cost DHFRP bar configurations while enforcing essential constraints. Applications for bridge decks and building beams are considered, with 2, 3, and 4-material bars. It was found that optimal bar configuration has little variation for the different applications, and that overall optimized bar cost decreased as the number of bar materials increased.

Reliability-Based Optimization Of Fiber-Reinforced Polymer Composite Bridge Deck Panels, Michel D. Thompson, Christopher D. Eamon, Masoud Rais-Rohani

Civil and Environmental Engineering Faculty Research Publications

A reliability-based optimization (RBO) procedure is developed and applied to minimize the weight of eight fiber-reinforced polymer composite bridge deck panel configurations. The method utilizes interlinked finite element, optimization, and reliability analysis procedures to solve the weight minimization problem with a deterministic strength constraint and two probabilistic deflection constraints. Panels are composed of an upper face plate, lower face plate, and a grid of interior stiffeners. Different panel depths and stiffener layouts are considered. Sensitivity analyses are conducted to identify significant design and random variables. Optimization design variables are panel component ply thicknesses while random variables include load and material …

Design Optimization Module For Hierarchical Research And Learning Environment, Oktay Baysal, Mehti Koklu, Ahmed K. Noor

Mechanical & Aerospace Engineering Faculty Publications

The present paper describes a learning module on design optimization courses within a hierarchical research and learning network (HRLN). In this environment a knowledge organization can be created as a hierarchical learning network to link diverse inter- and trans- disciplinary teams from a consortium of universities, industry, government agencies and the providers of learning technologies. It is an approach that builds on computer-based training, intelligent tutoring systems, interactive learning, collaborative-distributed learning, and learning networks. The present design optimization module has been developed and described herein, as a demonstrator of a learning module in this environment. This module allows for the …