Showing 1–2 of 2 results for author: Phillips, T
-
Multi-fidelity power flow solver
Authors:
Sam Yang,
Bjorn Vaagensmith,
Deepika Patra,
Ryan Hruska,
Tyler Phillips
Abstract:
We propose a multi-fidelity neural network (MFNN) tailored for rapid high-dimensional grid power flow simulations and contingency analysis with scarce high-fidelity contingency data. The proposed model comprises two networks -- the first one trained on DC approximation as low-fidelity data and coupled to a high-fidelity neural net trained on both low- and high-fidelity power flow data. Each networ…
▽ More
We propose a multi-fidelity neural network (MFNN) tailored for rapid high-dimensional grid power flow simulations and contingency analysis with scarce high-fidelity contingency data. The proposed model comprises two networks -- the first one trained on DC approximation as low-fidelity data and coupled to a high-fidelity neural net trained on both low- and high-fidelity power flow data. Each network features a latent module which parametrizes the model by a discrete grid topology vector for generalization (e.g., $n$ power lines with $k$ disconnections or contingencies, if any), and the targeted high-fidelity output is a weighted sum of linear and nonlinear functions. We tested the model on 14- and 118-bus test cases and evaluated its performance based on the $n-k$ power flow prediction accuracy with respect to imbalanced contingency data and high-to-low-fidelity sample ratio. The results presented herein demonstrate MFNN's potential and its limits with up to two orders of magnitude faster and more accurate power flow solutions than DC approximation.
△ Less
Submitted 26 May, 2022;
originally announced May 2022.
-
A Method of Predicting Powder Flowability for Selective Laser Sintering
Authors:
Douglas Sassaman,
Timothy Phillips,
Joseph J Beaman,
Craig Milroy,
Matthew Ide
Abstract:
This work investigates a method for pre-screening material systems for Selective Laser Sintering (SLS) using a combination of Revolution Powder Analysis (RPA) and machine learning. To develop this method, nylon was mixed with alumina or carbon fibers in different wt.% to form material systems with varying flowability. The materials were measured in a custom RPA device and the results compared with…
▽ More
This work investigates a method for pre-screening material systems for Selective Laser Sintering (SLS) using a combination of Revolution Powder Analysis (RPA) and machine learning. To develop this method, nylon was mixed with alumina or carbon fibers in different wt.% to form material systems with varying flowability. The materials were measured in a custom RPA device and the results compared with as-spread layer density and surface roughness. Machine learning was used to attempt classification of all powders for each method. Ultimately, it was found that the RPA method is able to reliably classify powders based on their flowability, but as-spread layer density and surface roughness were not able to be classified.
△ Less
Submitted 25 June, 2021;
originally announced June 2021.
