A Review on Real Driving Cycle-Based State of Charge Prediction for EV Batteries Ikhsan Romli, Bermawi Priyatna Iskandar
1 Teknik dan Manajemen Industri, Fakultas Teknologi Industri, Institut Teknologi Bandung, Indonesia,
2 Teknik Industri, Fakultas Teknik, Universitas Pelita Bangsa, Indonesia
Abstract
Research on performance of Electric Vehicle is very important, especially in driving range of a Battery Electric Vehicle (BEV) that requires precise State of Charge (SoC) predictions. The battery SoC is an important parameter that reflects the performance of the battery. Meanwhile, the battery has varying time properties depending on real conditions when driving. It has a logical relationship in a strong non-linear form that makes it very complex. Therefore, SoC prediction based on the Real Driving Cycle (RDC) can accurately protect the battery, save energy, increase battery life, prevent overcharging or discharging, and also enable applications to make rational control strategies to achieve goals with a certain range. This paper provides a literature review of various papers that are relevant and related to SoC prediction method for BEVs based on RDC. This paper summarizes the approaches used in Li-ion battery SoC prediction. Three approaches are classified accordingly, i.e. model-based approach, data-driven approach and hybrid approach. The results achieved imply that data-driven models, especially machine learning methods have the best accuracy. Based on the assessment of the various SOC prediction methods reviewed, the key issues and direction of developing SOC prediction in the future trend are also discussed.
Keywords: State of charge- real driving- prediction
