EV Battery State of Charge Estimation Using Machine Learning скачать в хорошем качестве

EV Battery State of Charge Estimation Using Machine Learning

By Talha Kamran, Andrew Chacko, and Nemesh Weerawarna at the Memorial University of Newfoundland, Canada. The State of Charge (SOC) performs the same function as the fuel gauge in a fossil fuel-powered vehicle, indicating how much energy remains inside a battery to power a vehicle. Properly measuring SOC is difficult since it cannot be measured directly due to non-linear, time-varying properties and electrochemical processes. Moreover, battery SOC is impacted by other factors such as age, temperature variations, charge-discharge cycles, and so on. This presentation showcases machine learning to estimate the State of Charge of an LG HG2 18650 lithium-ion cell using publicly available battery test data from McMaster University in Ontario, Canada. Several experiments are run against Deep Neural Networks (DNN), Convolutional Neural Networks (CNN), and Long Short-Term Memory Networks (LSTM). The prediction results from these are then compared to each other.