Towards Sustainable Consumer Electronics: DL-based SoH and RUL Prediction for E-Waste Reduction
Anureet Chhabra, Sarjana Singh, Akash Sharma, Sudhakar Kumar, Brij B. Gupta, Varsha Arya, Kwok Tai Chui
Abstract
Lithium-ion (Li-ion) batteries are gaining attention as they are crucial for powering a vast range of consumer electronics including smartphones, laptops, and Electric vehicles (EVs). However, if the State of Health (SoH) and Remaining Useful Life (RUL) of the battery is not monitored constantly, it can decrease the battery’s performance, and maximum capacity, which leads to a continual reduction in lifespan and a reduction in the driving range in case of EVs. Since SoH and RUL are very important factors in consumer electronics battery management systems accurate prediction and regular monitoring are required to achieve a longer lifetime of the battery which will help in reducing E-waste and achieving the Sustainable Development Goals (SDG-7): affordable and clean energy. This study presents a novel approach to predict the SoH and RUL of Li-ion batteries in consumer electronics using deep learning (DL), and LSTM networks. The proposed model is trained on a benchmark dataset by NASA for Li-ion Battery Aging. Experimental results presented in this study demonstrate the effectiveness of the DL-based approach in accurately predicting the SoH and RUL with RMSE values of 0.050681 and 0.04092 respectively.