Prediction of Lithium-Ion Battery Capacity in UAVs

摘要

This article proposes a research on degradation of lithium-ion batteries. The life of these batteries can be measured by monitoring capacity degradation. A dataset made available by NASA for the present study was used. Also, 3 methods were used to predict the failure. The methods were Non-recursive Least Squares, a Kalman Filter and an ELM neural network with a very simple architecture, but the results achieved good results and can be compared to other more robust algorithms.