Real-Time State-of-Charge Tracking Embedded in the Advanced Driver Assistance System of Electric Vehicles

Abstract

Range anxiety has remained a critical technological bottleneck in electric vehicles (EVs). The issue lies in the fact that the estimation of the EV's batteries state of charge (SoC) is still inaccurate and unreliable due to the complex and nonlinear characteristics of the batteries. To tackle the problem, this paper presents a novel real-time mixed SoC estimation algorithm for the EV's lithium-ion batteries for implementation in an advanced driver assistance system (ADAS). The mixed estimation algorithm combines: i) an improved Coulomb counting method (CCM) that takes into account the battery state-of-health, operating temperature, and aging effect, ii) a model-based method (MBM) that represents a real-time recursive structure of the battery, and iii) a bottom-up based method (BUBM) that takes into account a variety of the environmental conditions, traffic conditions, auxiliary loads, and driver's behavior. To validate the proposed algorithm, several laboratory tests under real-time driving cycles have been conducted on a Manganese-oxide Li-ion cell of a 2012 Nissan Leaf battery cell. Furthermore, the effectiveness of the model has been demonstrated by driving the Nissan Leaf along a selected route in Australia. The results demonstrate a great accuracy for the SoC estimation compared to previous models.