Propellant Mass Gauging in a Spherical Tank under Micro-Gravity Conditions Using Capacitance Plate Arrays and Machine Learning

Abstract

Propellant mass gauging under micro-gravity conditions is a challenging task due to the unpredictable position and shape of the fuel body inside the tank. Micro-gravity conditions are common for orbiting satellites and rockets that operate on limited fuel supplies. Capacitance sensors have been investigated for this task in recent years; however, the effect of various positions and shapes of the fuel body is not analyzed in detail. In this paper, we investigate this with various fill types, such as annular, core-annular, and stratified fills at different positions. We compare the performance among several curve-fitting-based approaches and a machine-learning-based approach, the latter of which offers superior performance in estimating the fuel content.

Keywords

• capacitance sensors
• electrical capacitance volume tomography
• two-phase flow
• micro-gravity mass gauging
• machine learning