155 / 2023-09-30 21:40:56

A DRT based Lithium battery multi-time scale model parameter identification and SOC estimation

multi-time scale, DRT, state of charge, EIS

Special Sessions > SS03 - Sensor related circuit and system, and Artificial Intelligence Algorithm

Whether in the field of energy storage or electric vehicle applications, the key function of the lithium battery

management system is to calculate the accurate charging state online in real time through the detected voltage, current and temperature, which directly depends on the accuracy of the parameter identification in the equivalent circuit model (ECM). In this paper, a new online multi-time scale estimation method is proposed to determine the structure of RC network in ECM of Li-ion batteries, and identify model parameters at different time scales. Firstly, the dynamic characteristics of battery system dominated by different electrochemical effects at different time scales were analyzed by EIS test. The distribution of relaxation time (DRT) method is used to determine the ECM structure of the battery by identifying the time constants representing different electrochemical processes, and to distinguish the time scale peaks. At the same time, the parameters of ECM model on different time scales are decoupled to overcome the numerical

problems of the classical recursive least squares method in the identification process. In order to update open circuit voltage (OCV) online in real time, OCV is used as a slow time scale parameter for collaborative online estimation. Through the Urban Dynamometer Driving Schedule (UDDS) experiment, the classical forgetting factor recursive least squares (FFRLS) and the proposed multi-time scale iterative least squares (MTRLS) method without SOC-OCV on-line experiment are implemented. The identification results of ECM parameters and SOC under the two methods are compared and discussed. The results show that compared with FFRLS technique, the SOC identification accuracy of the proposed estimation method are improved by

1.65%