Prediction of Retained Capacity and EODV of Li-ion Batteries in LEO Spacecraft Batteries

TL;DR

This paper develops an ANN model to predict the retained capacity and EODV of Li-ion batteries in LEO spacecraft, considering 25000 cycles under various conditions, validated against experimental data.

Contribution

It introduces a novel ANN modeling approach for long-cycle Li-ion batteries in space applications, incorporating temperature and DOD variations.

Findings

ANN accurately predicts battery capacity and EODV over 25000 cycles.

Model validation shows strong agreement with experimental results.

The approach aids in reliable battery life estimation for space missions.

Abstract

In resent years ANN is widely reported for modeling in different areas of science including electro chemistry. This includes modeling of different technological batteries such as lead acid battery, Nickel cadmium batteries etc. Lithium ion batteries are advance battery technology which satisfy most of the space mission requirements. Low earth orbit (LEO)space craft batteries undergo large number of charge discharge cycles (about 25000 cycles)compared to other ground level or space applications. This study is indented to develop ANN model for about 25000 cycles, cycled under various temperature, Depth Of Discharge (DOD) settings with constant charge voltage limit to predict the retained capacity and End of Discharge Voltage (EODV). To extract firm conclusion and distinguish the capability of ANN method, the predicted values are compared with experimental result by statistical method and Bland Altman plot.