NASA Logo

NTRS

NTRS - NASA Technical Reports Server

Back to Results
Machine Learning for Dynamical Modeling of a Flexible Inverted Pendulum SystemThe inverted pendulum system, a canonical example of an unstable mechanical system, is often used to model the control problems encountered in the flight of rockets in the initial stages of launch, when the airspeed is too small for aerodynamic stability. A system with a flexible pendulum is a variant that more accurately simulates rocket flight nonlinearities (particularly, the flex modes of the rocket). To increase NASA capability of modeling dynamical systems for which closed form solutions are not clear or easily developed, this project aims to provide a machine learning approach that produces a learned dynamical model of the PENNY robot (a rover with a flexible inverted pendulum) from operational data. The developed approach can then be generalized to other complex dynamical systems, including but not limited to rockets and other robotic systems.
Document ID
20205004568
Document Type
Other - Abstract
Authors
M. A. DuPuis (Kennedy Space Center Merritt Island, Florida, United States)
N. A. Janmohamed (Santa Monica College Santa Monica, California, United States)
Date Acquired
July 16, 2020
Publication Date
July 21, 2020
Subject Category
Cybernetics, Artificial Intelligence and Robotics
Funding Number(s)
CONTRACT_GRANT: NNX13AJ45A
Distribution Limits
Public
Copyright
Use by or on behalf of the US Gov. Permitted.
Technical Review
NASA Technical Management

Available Downloads

NameType Abstract-Nashir-Janmohamed-Summer-20.pdf STI