Adaptive model-based state monitoring algorithms for lithium-ion batteries

  • Adaptive modellbasierte Zustandsüberwachungsalgorithmen für Lithium-Ionen-Batterien

Li, Shi; Pischinger, Stefan (Thesis advisor); Andert, Jakob Lukas (Thesis advisor)

Aachen (2020, 2021)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2020


Lithium-ion batteries are the prevalent technology for the state-of-the-art energy storage system in electric vehicles (EV). Battery management system (BMS) is used to guarantee the safe and efficient operation of the battery system. One of the core functions of BMS is to monitor the internal states of the battery such as the state of charge (SOC) and the state of health (SOH). This work investigates the advanced model-based algorithms for battery state monitoring with each step evaluated and elaborated using literature study, simulative implementation, comparative study, and verification. First, an experimental protocol is built, with which the automotive battery cell is characterized. Second, a model is selected and parameterized based on the measured data. With proper model obtained, techniques from the control theory including the extended Kalman filter (EKF), the particle filter (PF), and the recursive least square method (RLS) are implemented for the SOC and SOH estimations. Data obtained from different working conditions and the accelerated aging test are used in a model-in-the-loop (MIL) environment for algorithm verifications. The parameterization of the filters is identified to profoundly influence the estimation results. A novel method utilizing the learning ability of the adaptive neuro-fuzzy inference system (ANFIS) is proposed to update the noise covariance matrixes of the filters online. The proposed method demonstrates promising accuracies with the root-mean-square error smaller than 2% and improved robustness in battery state estimations under different operating conditions. This would reduce the effort of filter tuning, further allow more efficient monitoring and more optimal sizing of the battery system.