2020/12/19 by Aniruddh Herle, Herle, Aniruddh, Janamejaya Channegowda +3
Engineering · #Advanced Battery Technologies Research #Electric Vehicles and Infrastructure #Electric and Hybrid Vehicle Technologies #FOS: Computer and information sciences #FOS: Electrical engineering #Machine Learning (cs.LG) #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2012.10725
openalex publication_date 2020/12/19 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
Electric Vehicles (EVs) are rapidly increasing in popularity as they are\nenvironment friendly. Lithium Ion batteries are at the heart of EV technology\nand contribute to most of the weight and cost of an EV. State of Charge (SOC)\nis a very important metric which helps to predict the range of an EV. There is\na need to accurately estimate available battery capacity in a battery pack such\nthat the available range in a vehicle can be determined. There are various\ntechniques available to estimate SOC. In this paper, a data driven approach is\nselected and a Nonlinear Autoregressive Network with Exogenous Inputs Neural\nNetwork (NARXNN) is explored to accurately estimate SOC. NARXNN has been shown\nto be superior to conventional Machine Learning techniques available in the\nliterature. The NARXNN model is developed and tested on various EV Drive Cycles\nlike LA92, US06, UDDS and HWFET to test its performance on real world\nscenarios. The model is shown to outperform conventional statistical machine\nlearning methods and achieve a Mean Squared Error (MSE) in the 1e-5 range.\n