A novel method for SOC estimation of Li-ion batteries using a hybrid machine learning technique


Ipek E., Yılmaz M.

TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES, vol.29, no.1, pp.18-31, 2021 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 29 Issue: 1
  • Publication Date: 2021
  • Doi Number: 10.3906/elk-1912-42
  • Title of Journal : TURKISH JOURNAL OF ELECTRICAL ENGINEERING AND COMPUTER SCIENCES
  • Page Numbers: pp.18-31

Abstract

The battery system is one of the key components of electric vehicles (EV) which has brought groundbreaking technologies. Since modern EVs have mostly Li-ion batteries, they need to be monitored and controlled to achieve safe and high-performance operation. Particularly, the battery management system (BMS) uses complex processing systems that perform measurements, estimation of the battery states, and protection of the system. State of charge (SOC) estimation is a major part of these processes which defines remaining capacity in the battery until the next charging operation as a proportion to the total battery capacity. Since SOC is not a parameter that can be measured, the fundamental challenge is an accurate estimation. There are different SOC estimation methods in the literature that promises high accuracy such as model-based estimations, adaptive filter based estimations, and a combination of these systems. Recently, artificial intelligence (AI) and particularly machine learning (ML) based systems are included in the battery state estimation both as a part of adaptive systems and standalone. Data-driven methods are promising approaches to battery state estimation which provide high accuracy. The purpose of this study is to present a novel and highly accurate way of SOC estimation of the Li-ion battery (LIB) cell with a considerably low parameterization and modeling effort. Therefore, a new approach is proposed to estimate SOC with reduced modeling and without performing parametrization. Based on discharge test data, XGBoost is used to estimate SOC under dynamic operating conditions and the estimation is reached 98.81% coefficient of determination. As a novel approach, exponential smoothing is performed in combination with XGBoost SOC estimation to improve the estimation performance of the model. The estimation accuracy is improved as approximately 0.62%.