This project utilizes machine learning methods to examine how subject-specific differences influence hand function and create subject-specific computer models from easy to obtain clinical data. Completion of this project will critically advance the ability to efficiently create subject-specific models of the hand and understand the biomechanical mechanism underlying hand force production.
Tag: Joel Harley
Spread Spectrum Time Domain Reflectivity for String Monitoring in PV Power Plants
This project is investigating the application of Spread Spectrum Time Domain Reflectivity (SSTDR) to monitor operating strings of modules in large photovoltaic (PV) arrays. SSTDR can detect, and spatially localize, changes in the impedance of the system in real-time, including at high voltages and currents. This allows monitoring of intermittent and slowly-evolving degradation and failure modes, and potentially enables more efficient characterization of PV power plants, which will maximize future energy output, reduce the levelized cost of electricity, and increase bankability.
EAGER: Real-Time: Ultrasonic Reconstruction and Localization with Deep Helmholtz Networks
The objective of the project is to establish the foundation for Helmholtz networks, which are deep, generative, physics-informed neural networks that reconstruct ultrasonic wave propagation and locate ultrasonic sources. The Helmholtz networks are based on the fact that each frequency of a wave can be represented as the sum of a sparse number of spatial modes. The modes are constrained by the Helmholtz equation and this physical constraint ensures that the machine learning algorithm is trustworthy for system-critical engineered systems (e.g., health monitoring of an aircraft). Such physics-informed machine learning is an important (albeit not widely studied) topic for integrating advanced computation tools into real-time engineered systems.
Elucidating Grain Growth in Thermo-Magnetic Processed Materials by Transfer Learning and Reinforcement Learning
The goal of the proposed work is to combine deep, model-based reinforcement learning and transfer learning to elucidate one of the most fundamental, yet poorly understood, mechanisms in materials science: abnormal grain growth. We hypothesize that abnormal grain growth is the result of highly anisotropic grain boundary character networks, where a unique combination of neighboring grain boundaries incentivize accelerated growth.