Electric Vehicle Battery Pack Prediction of Capacity Degradation Based on Deep Learning Architecture and Internet of Things

As electric cars (EVs) become more widespread, research into battery life is becoming highly significant. Electric car batteries are now constantly connected and transmit information on a massive scale. The huge amounts of data produced, quite apart from the widespread use of the internet by these battery packs, provide new challenges for researchers and regulators. A unique deep learningDeep learning model with use of internet of thingsInternet of things (IoT) is suggested in this paper to produce a universal and accurate Li-ion battery aging prediction. On the other hand, deep learning (DL) is an efficient strategy for handling IoT problems including data analysis, prediction, and categorization. However, it is challenging to get the best data for deep learningDeep learning in IoT for real-time prediction. The accuracy and robustness of prediction will be determined by data collecting components such as sensors and cameras. In this article, deep learningDeep learning model Long Short-Term Memory (LSTM) method is used to training and testing of the deep learningDeep learning architecture. Root Mean Square Error (RMSE) value for the deep learningDeep learning model is 0.69, which will accurately predict vehicle battery pack data. Numerous articles have been published on improving deep learningDeep learning models for RUL estimate of battery packs; however, there is no research of capacity degradation of battery pack estimation using IoT and deep learningDeep learning approaches in the literature. In this research, a strategy for estimating the RUL of an electric car battery pack is described using deep learningDeep learning and IoT.