State of health (SOH) prediction for Lithium-ion batteries using regression and LSTM

Sunwoda Electronic Co., Ltd, and Tsinghua Berkeley Shenzhen Institute (TBSI) generate the TBSI Sunwoda Battery Dataset. We open-source this dataset to inspire more data-driven novel material verification, battery management research and applications.

An awesome list of papers on remaining useful life (RUL) prediction from arXiv

A comprehensive simulation platform integrating vehicle dynamics, environment emulation, body controls, and battery management for holistic testing and validation of automated vehicles.

Repository of my master’s thesis "Development and evaluation of a model for predicting the state of health of traction batteries based on artificial neural networks"

translation for paper Machine learning pipeline for battery state-of-health estimation

Code and models for estimating the State of Charge (SoC) and of battery cells. Utilizing advanced deep learning techniques.

Sandbox to develop, test and compare Kalman Fitler-enabled estimation techniques for state of charge of a sample lithium-ion battery, utilizing transient signals to predict state across points in time.

The project analyzes battery cycling data to predict degradation patterns and performance metrics using both deep learning (LSTM) and traditional machine learning (XGBoost) approaches. The implementation enables accurate estimation of battery health, which is crucial for battery management systems in various applications.

Master Thesis in Data Science and Engineering