Publication: Sparse Regression for the Breathing Mode Instability: Extracting Governing Equations from Hall Effect Thrusters Simulation Data
Loading...
Identifiers
Publication date
2023
Defense date
2023-07-06
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
The development of efficient and reliable electric space propulsion systems relies on accurate modeling and identification of their underlying dynamics. Traditional approaches to model identification often involve intricate physical analysis or making extensive assumptions, limiting their applicability and scalability. In this thesis an algorithm for accurate modeling and identification of electric space propulsion systems is presented. The algorithm, based on sparse regression and model parsimony, allows automatic data-driven identification of models for space plasma thrusters. It incorporates statistical techniques, physical constraints, and trajectory-based information for robust system identification. The algorithm is demonstrated using PIC/fluid simulation data from a Hall Effect Thruster for several operating points. Models of varying complexity are obtained, focusing on physical explainability and coefficient variation with operating point. The resulting equations for average ion and neutral densities align well with existing models. Point-wise density models exhibit location dependency in the discharge chamber. The algorithm showcases its general applicability to other electric propulsion systems.